科技论文题目The Application of LCD系别尚德光伏学院专业微电子技术班级1001学生姓名于路学号100100193指导教师谢丽君2013年4 月On the deployment of VolP in Ethernet networks:methodology and case studyKhaled Salah, Department of Information and Computer Science, King Fahd University of Petroleum and Minerals, P.O. Box 5066, Dhahran 31261, Saudi ArabiaReceived 25 May 2004; revised 3 June 2005; accepted 3 June 2005. Available online 1 July 2005.AbstractDeploying IP telephony or voice over IP (V olP) is a major and challenging task for data network researchers and designers. This paper outlines guidelines and a step-by-step methodology on how V olP can be deployed successfully. The methodology can be used to assess the support and readiness of an existing network. Prior to the purchase and deployment of V olP equipment, the methodology predicts the number of V olP calls that can be sustained by an existing network while satisfying QoS requirements of all network services and leaving adequate capacity for future growth. As a case study, we apply the methodology steps on a typical network of a small enterprise. We utilize both analysis and simulation to investigate throughput and delay bounds. Our analysis is based on queuing theory, and OPNET is used for simulation. Results obtained from analysis and simulation are in line and give a close match. In addition, the paper discusses many design and engineering issues. These issues include characteristics of V olP traffic and QoS requirements, V olP flow and call distribution, defining future growth capacity, and measurement and impact of background traffic.Keywords: Network Design，Network Management，V olP，Performance Evaluation，Analysis，Simulation，OPNET1 IntroductionThese days a massive deployment of V olP is taking place over data networks. Most ofthese networks are Ethernet based and running IP protocol. Many network managers are finding it very attractive and cost effective to merge and unify voice and data networks into one. It is easier to run, manage, and maintain. However, one has to keep in mind that IP networks are best-effort networks that were designed for non-real time applications. On the other hand, V olP requires timely packet delivery with low latency, jitter, packet loss, and sufficient bandwidth. To achieve this goal, an efficient deployment of V olP must ensure these real-time traffic requirements can be guaranteed over new or existing IP networks. When deploying a new network service such as V olP over existing network, many network architects, managers, planners, designers, and engineers are faced with common strategic, and sometimes challenging, questions. What are the QoS requirements for V olP? How will the new V olP load impact the QoS for currently running network services and applications? Will my existing network support V olP and satisfy the standardized QoS requirements? If so, how many V olP calls can the network support before upgrading prematurely any part of the existing network hardware? These challenging questions have led to the development of some commercial tools for testing the performance of multimedia applications in data networks. A list of the available commercial tools that support V olP is listed in [1,2]. For the most part, these tools use two common approaches in assessing the deployment of V olP into the existing network. One approach is based on first performing network measurements and then predicting the network readiness for supporting V olP. The prediction of the network readiness is based on assessing the health of network elements. The second approach is based on injecting real V olP traffic into existing network and measuring the resulting delay, jitter, and loss. Other than the cost associated with the commercial tools, none of the commercial tools offer a comprehensive approach for successful V olP deployment. I n particular, none gives any prediction for the total number of calls that can be supported by the network taking into account important design and engineering factors. These factors include V olP flow and call distribution, future growth capacity, performance thresholds, impact of V olP on existing network services and applications, and impact background traffic on V olP. This paper attempts to address those important factors and layout a comprehensive methodology for a successful deployment of any multimedia application such as V olP and video conferencing. However, thepaper focuses on V olP as the new service of interest to be deployed. The paper also contains many useful engineering and design guidelines, and discusses many practical issues pertaining to the deployment of V olP. These issues include characteristics of V olP traffic and QoS requirements, V olP flow and call distribution, defining future growth capacity, and measurement and impact of background traffic. As a case study, we illustrate how our approach and guidelines can be applied to a typical network of a small enterprise. The rest of the paper is organized as follows. Section 2 presents a typical network topology of a small enterprise to be used as a case study for deploying V olP. Section 3 outlines practical eight-step methodology to deploy successfully V olP in data networks. Each step is described in considerable detail. Section 4 describes important design and engineering decisions to be made based on the analytic and simulation studies. Section 5 concludes the study and identifies future work.Moreover, industrial automation communication protocols have not reached the same level of standardization as office communication networks, which further justifies the predominance of proprietary architectures.However, the success of the Internet and of the Web has started impacting the industrial V olP world too. Industrial users are starting to familiarize with Web interfaces, graphical quality, multimedia content, and features such as mobility, adaptivity, and personalization of the applications.At the same time,TCP-IP based communication protocols and embedded operating systems have started to spread in the industrial automation field [6][10], thus reducing the need of proprietary architectures making enterprisewide integration more appealing.In this scenario, it is easy to foresee a slow but inexorable convergence of the industrial V olP solutions towards standard architectures, standard communication protocols, and advanced interactive functions.Our work focuses on the design of a new distributed software architecture for VOLP systems able to provide features and services such as personalization, adaptivity, distribution, mobility, multi-channel notification, integration with office networks and software packages, although preserving the robustness, reliability, performance and cost-effectiveness oftraditional V olP solutions.The project, called ESAMyV olP is a research activity carried out in collaboration between Politecnico di Milano and ESA Elettronica S.p.A., an Italian company operating in the V olP market.2 Existing networkFig. 1 illustrates a typical network topology for a small enterprise residing in a high-rise building. The network shown is realistic and used as a case study only; however, our work presented in this paper can be adopted easily for larger and general networks by following the same principles, guidelines, and concepts laid out in this paper. The network is Ethernet-based and has two Layer-2 Ethernet switches connected by a router. The router is Cisco 2621, and the switches are 3Com Superstack 3300. Switch 1 connects Floors 1 and 2 and two servers; while Switch 2 connects Floor 3 and four servers. Each floor LAN is basically a shared Ethernet connecting employee PCs with workgroup and printer servers. The network makes use of VLANs in order to isolate broadcast and multicast traffic. A total of five LANs exist. All VLANs are port based. Switch 1 is configured such that it has three VLANs. VLAN1 includes the database and file servers. VLAN2 includes Floor 1. VLAN3 includes Floor2. On the other hand, Switch 2 is configured to have two VLANs. VLAN4 includes the servers for E-mail, HTTP, Web and cache proxy, and firewall. VLAN5 includes Floor 3. All the links are switched Ethernet 100 Mbps full duplex except for the links for Floors 1–3 which are shared Ethernet 100 Mbps half duplex.Industrial V olP products rarely implement innovative services, such as remote access to the plant control, messaging and remote notification.Indeed, V olP companies seem to privilege exclusively performance and good access to industrial communication standards, even if these factors could be incompatible with the adoption of innovative solution based on modern and solid Web architectures.Even the V olP players that seem to offer the most innovative contents (and claim their products are Web-enabled) still leverage on legacy architectures, typically exploiting monolithic Applications.On the contrary, recent studies [6] show how users are increasingly looking towards a new range of products with advanced features, superior graphical capabilities and improved usability that could grant:• remote and possibly, distributed control of an industrial plant;• remote notification solutions even when the user is not in front of the terminal;• personalization and automatic adaptation of the GUI;• integration with existing enterprise processes, systems and equipments;• openness to new standard and best practises in the field, by offering low cost modularity and extensibility.SCADA (Supervisory Control And Data Acquisition) systems recently introduced some interesting innovations but, as the acronym suggests, their target is focused on products that implement a wide range of high-level functionalities and that can be deployed in a large set of contexts.They are typically deployed on high-profile devices (PCs and powerful embedded systems) and represent a niche in the V olP market.In the other market’s sectors, innovation has been led by main vendors (e.g., Siemens), who have been working for the past few years in raising the level of the features provided by traditional V olP applications.Sm@rtAccess [15], for example, is a technology developed that allows distributing the control of an industrial plant over a maximum of three stations.Its functioning, though, is based on simply broadcasting the displayed interface of the apparatus that is directly connected with the plant to the others clients.The bandwidth requirements of this approach exceed the capability of a typical Internet connection.Progea [16] proposes a more innovative solution by offering remotization features and a Web-based architecture.Running the Progea server application on a Windows XP based PC, it is possible to remotely control a plant from an internet connected standard Web browser that has the support of a JVM (Java Virtual Machine).Even if powerful, this approach lacks in offering a portable solution since differentimplementations have been provided for different platforms.3 Step-by-step methodologyFig. 2 shows a flowchart of a methodology of eight steps for a successful V olP deployment. The first four steps are independent and can be performed in parallel. Before embarking on the analysis and simulation study, in Steps 6 and 7, Step 5 must be carried out which requires any early and necessary redimensioning or modifications to the existing network. As shown, both Steps 6 and 7 can be done in parallel. The final step is pilot deployment.3.1. V olP traffic characteristics, requirements, and assumptionsFor introducing a new network service such as V olP, one has to characterize first the nature of its traffic, QoS requirements, and any additional components or devices. For simplicity, we assume a point-to-point conversation for all V olP calls with no call conferencing. For deploying V olP, a gatekeeper or Call Manager node has to be added to the network [3,4,5]. The gatekeeper node handles signaling for establishing, terminating, and authorizing connections of all V olP calls. Also a V olP gateway is required to handle external calls. A V olP gateway is responsible for converting V olP calls to/from the Public Switched Telephone Network (PSTN). As an engineering and design issue, the placement of these nodes in the network becomes crucial. We will tackle this issue in design step 5. Other hardware requirements include a V olP client terminal, which can be a separate V olP device, i.e. IP phones, or a typical PC or workstation that is V olP-enabled. A V olP-enabled workstation runs V olP software such as IP Soft Phones .Fig. 3 identifies the end-to-end V olP components from sender to receiver [9]. The first component is the encoder which periodically samples the original voice signal and assigns a fixed number of bits to each sample, creating a constant bit rate stream. The traditionalsample-based encoder G.711 uses Pulse Code Modulation (PCM) to generate 8-bit samples every 0.125 ms, leading to a data rate of 64 kbps . The packetizer follows the encoder and encapsulates a certain number of speech samples into packets and adds the RTP, UDP, IP, and Ethernet headers. The voice packets travel through the data network. An important component at the receiving end, is the playback buffer whose purpose is to absorb variations or jitter in delay and provide a smooth playout. Then packets are delivered to the depacketizer and eventually to the decoder which reconstructs the original voice signal. We will follow the widely adopted recommendations of H.323, G.711, and G.714 standards for V olP QoS requirements.Table 1 compares some commonly used ITU-T standard codecs and the amount of one-way delay that they impose. To account for upper limits and to meet desirable quality requirement according to ITU recommendation P.800, we will adopt G.711u codec standards for the required delay and bandwidth. G.711u yields around 4.4 MOS rating. MOS, Mean Opinion Score, is a commonly used V olP performance metric given in a scale of 1–5, with 5 is the best. However, with little compromise to quality, it is possible to implement different ITU-T codecs that yield much less required bandwidth per call and relatively a bit higher, but acceptable, end-to-end delay. This can be accomplished by applying compression, silence suppression, packet loss concealment, queue management techniques, and encapsulating more than one voice packet into a single Ethernet frame.3.1.1. End-to-end delay for a single voice packetFig. 3 illustrates the sources of delay for a typical voice packet. The end-to-end delay is sometimes referred to by M2E or Mouth-to-Ear delay. G.714 imposes a maximum total one-way packet delay of 150 ms end-to-end for V olP applications . In [22], a delay of up to 200 ms was considered to be acceptable. We can break this delay down into at least three different contributing components, which are as follows (i) encoding, compression, and packetization delay at the sender (ii) propagation, transmission and queuing delay in the network and (iii) buffering, decompression, depacketization, decoding, and playback delay atthe receiver.3.1.2. Bandwidth for a single callThe required bandwidth for a single call, one direction, is 64 kbps. G.711 codec samples 20 ms of voice per packet. Therefore, 50 such packets need to be transmitted per second. Each packet contains 160 voice samples in order to give 8000 samples per second. Each packet is sent in one Ethernet frame. With every packet of size 160 bytes, headers of additional protocol layers are added. These headers include RTP+UDP+IP+Ethernet with preamble of sizes 12+8+20+26, respectively. Therefore, a total of 226 bytes, or 1808 bits, needs to be transmitted 50 times per second, or 90.4 kbps, in one direction. For both directions, the required bandwidth for a single call is 100 pps or 180.8 kbps assuming a symmetric flow.3.1.3. Other assumptionsThroughout our analysis and work, we assume voice calls are symmetric and no voice conferencing is implemented. We also ignore the signaling traffic generated by the gatekeeper. We base our analysis and design on the worst-case scenario for V olP call traffic. The signaling traffic involving the gatekeeper is mostly generated prior to the establishment of the voice call and when the call is finished. This traffic is relatively small compared to the actual voice call traffic. In general, the gatekeeper generates no or very limited signaling traffic throughout the duration of the V olP call for an already established on-going call. In this paper, we will implement no QoS mechanisms that can enhance the quality of packet delivery in IP networks.A myriad of QoS standards are available and can be enabled for network elements. QoS standards may include IEEE 802.1p/Q, the IETF’s RSVP, and DiffServ.Analysis of implementation cost, complexity, management, and benefit must be weighed carefully before adopting such QoS standards. These standards can be recommended when the cost for upgrading some network elements is high and the network resources are scarce and heavily loaded.3.2. V olP traffic flow and call distributionKnowing the current telephone call usage or volume of the enterprise is an importantstep for a successful V olP deployment. Before embarking on further analysis or planning phases for a V olP deployment, collecting statistics about of the present call volume and profiles is essential. Sources of such information are organization’s PBX, telephone records and bills. Key characteristics of existing calls can include the number of calls, number of concurrent calls, time, duration, etc. It is important to determine the locations of the call endpoints, i.e. the sources and destinations, as well as their corresponding path or flow. This will aid in identifying the call distribution and the calls made internally or externally. Call distribution must include percentage of calls within and outside of a floor, building, department, or organization. As a good capacity planning measure, it is recommended to base the V olP call distribution on the busy hour traffic of phone calls for the busiest day of a week or a month. This will ensure support of the calls at all times with high QoS for all V olP calls. When such current statistics are combined with the projected extra calls, we can predict the worst-case V olP traffic load to be introduced to the existing network.Fig. 4 describes the call distribution for the enterprise under study based on the worst busy hour and the projected future growth of V olP calls. In the figure, the call distribution is described as a probability tree. It is also possible to describe it as a probability matrix. Some important observations can be made about the voice traffic flow for inter-floor and external calls. For all these type of calls, the voice traffic has to be always routed through the router. This is so because Switchs 1 and 2 are layer 2 switches with VLANs configuration. One can observe that the traffic flow for inter-floor calls between Floors 1 and 2 imposes twice the load on Switch 1, as the traffic has to pass through the switch to the router and back to the switch again. Similarly, Switch 2 experiences twice the load for external calls from/to Floor 3.3.3. Define performance thresholds and growth capacityIn this step, we define the network performance thresholds or operational points for a number of important key network elements. These thresholds are to be considered when deploying the new service. The benefit is twofold. First, the requirements of the new service to be deployed are satisfied. Second, adding the new service leaves the network healthy andsusceptible to future growth. Two important performance criteria are to be taken into account. First is the maximum tolerable end-to-end delay; and second is the utilization bounds or thresholds of network resources. The maximum tolerable end-to-end delay is determined by the most sensitive application to run on the network. In our case, it is 150 ms end-to-end for V olP. It is imperative to note that if the network has certain delay sensitive applications, the delay for these applications should be monitored, when introducing V olP traffic, such that they do not exceed their required maximum values. As for the utilization bounds for network resources, such bounds or thresholds are determined by factors such as current utilization, future plans, and foreseen growth of the network. Proper resource and capacity planning is crucial. Savvy network engineers must deploy new services with scalability in mind, and ascertain that the network will yield acceptable performance under heavy and peak loads, with no packet loss. V olP requires almost no packet loss. In literature, 0.1–5% packet loss was generally asserted. However, in [24] the required V olP packet loss was conservatively suggested to be less than 105 . A more practical packet loss, based on experimentation, of below 1% was required in [22]. Hence, it is extremely important not to utilize fully the network resources. As rule-of-thumb guideline for switched fast full-duplex Ethernet, the average utilization limit of links should be 190%, and for switched shared fast Ethernet, the average limit of links should be 85% [25]. The projected growth in users, network services, business, etc. must be all taken into consideration to extrapolate the required growth capacity or the future growth factor. In our study, we will ascertain that 25% of the available network capacity is reserved for future growth and expansion. For simplicity, we will apply this evenly to all network resources of the router, switches, and switched-Ethernet links. However, keep in mind this percentage in practice can be variable for each network resource and may depend on the current utilization and the required growth capacity. In our methodology, the reservation of this utilization of network resources is done upfront, before deploying the new service, and only the left-over capacity is used for investigating the network support of the new service to be deployed.3.4. Perform network measurementsIn order to characterize the existing network traffic load, utilization, and flow, network measurements have to be performed. This is a crucial step as it can potentially affect results to be used in analytical study and simulation. There are a number of tools available commercially and noncommercially to perform network measurements. Popular open-source measurement tools include MRTG, STG, SNMPUtil, and GetIF [26]. A few examples of popular commercially measurement tools include HP OpenView, Cisco Netflow, Lucent VitalSuite, Patrol DashBoard, Omegon NetAlly, Avaya ExamiNet, NetIQ Vivinet Assessor, etc. Network measurements must be performed for network elements such as routers, switches, and links. Numerous types of measurements and statistics can be obtained using measurement tools. As a minimum, traffic rates in bits per second (bps) and packets per second (pps) must be measured for links directly connected to routers and switches. To get adequate assessment, network measurements have to be taken over a long period of time, at least 24-h period. Sometimes it is desirable to take measurements over several days or a week. One has to consider the worst-case scenario for network load or utilization in order to ensure good QoS at all times including peak hours. The peak hour is different from one network to another and it depends totally on the nature of business and the services provided by the network.The market of industrial V olP is seeing a slow but steady evolution towards the integration of industrial automation terminals with software and hardware architectures typical of office and Web-based applications, to achieve greater usability and flexibility of the interface and easier interoperability between industrial automation solutions and enterprise information systems.This goal requires unbundling the functions and modules of a traditional V olP solution, deploying them over a modular and distributed system, which exploits the open standards of the Internet and the architectural patterns of multi-tier Web Applications. The MyV olP project aims at designing, implementing and evaluating a distributed V olP platform which can be seamlessly accessed both locally and remotely and can be easily integrated in the enterprise ICT infrastructure. The major functional and non-functional requirements at the base of the MyV olP design are summarized in Table 1 and Table 2, respectively.Table 2 shows a summary of peak-hour utilization for traffic of links in both directions connected to the router and the two switches of the network topology of Fig. 1. These measured results will be used in our analysis and simulation study.4 The VolP architectureIn this section we overview the main characteristics of the design of the V olP framework. The overall architecture of the V olP platform is illustrated in Figure 1: the V olP functionality, usually embedded within the terminal attached to the controlled system, becomes partitioned into a client-server architecture, implemented on top of a hybrid communication network, comprising an Ethernet backbone that connects the V olP devices and a set of field bus protocols for connecting to the controlled plant.4.1. General design choicesThe design of the system had to address several issues, according to the requirements. In this section we give a summary of the adopted solutions.4.1.1. Distribution model of presentation andbusiness logic.The architecture design has been addressed by applying the state of the art solutions for granting separation of concerns and modular implementation. We adopted a Rich web interface paradigm consisting of extending the client side of traditional Web architectures, thus moving some of the computation from the server to the client.The business layer is still located at server side and contains the control policies, while the presentation layer is implemented at the client-side. It is responsible of building the interface for the human supervisor and of managing the user interaction.4.1.2. Personalization solutions.One of the most challenging requirements advocated for strong personalization capabilities of the platform. We use personalization based on groups that assumes users can be classified in roles, factoring out most of the personalization rules at the group level. The remaining detailed personalization rules can be applied on the single user, but we may assumethat their number and complexity are limited. This solution implements a good trade-off between the needs of fine-grained personalization and of reducing the complexity of the computation.4.1.3. Connectivity.The communication between client and server was implemented by means of HTTP with (emulated) callback through the HTTP request-response cycle. The client submits requests upon user interaction and upon timeouts generated by its internal clock. Requests submitted to the server are left pending until an actual update on the status of the controlled system happens. In this case, the server sends a response to the client, thus simulating an event-based message exchange.4.1.4. Personalization enactment.Some specific decisions must be taken about how and where to apply the personalization and adaptation rules on the Interfaces.Personalization and adaptivity rules could be stored and managed with two approaches:1.Rules encoded as XML files: personalization rules are generated by the offline configuration tool in XML format. Such rules are parsed and interpreted at runtime by a general purpose code, that generates the expected interface. A variant of the approach could devise several specific components for parsing the personalization and adaptivity rules addressing different issues (e.g., user interface, alarm configuration, and so on);2.Rules embedded in the code: this solution consists in generating and compiling a source code at configuration time that include specific personalization and adaptation rules targeted to the developed project. The result is a binary code that embeds all the rules and can be executed extremely fast, because no access to files or rule repositories is needed; We adopt both client and server side rule calculations: a hybrid approach in which part of the rules are applied at server side and the remaining ones at client side. We apply at client side the rules that affects the user interface and, in general, the client-side issues. For this part, the rules have been stored as binary code in the client application, for performance reasons. Vice versa, we adopted server-side application of the rules concerning the server configuration. In this case, the rules have been encoded as XML files and parsed by the server components.。
Research ProposalApplicant’s name: Huang TianqiTitlePublic Opinion Analysis Knowledge Enhanced Financial Decision Support in Big Data Environment1.AbstractThe big data era has brought great challenges to public opinion analysis and its application to financial decision support systems (FDSS). Under the environment, the original public opinion methods and its decision support module became insufficient for the application requirements. New public opinion analysis methods needs to be able to analyze not only all critical financial market information but also massive user-generated content, including blogs, tweets and posts. More importantly, FDSSs need to extend its knowledge base to new public opinion analysis so as to provide more valuable evidence for decision making. The aims of this study is to theoretically investigate the feasibility of adopting public opinion knowledge, provided by public opinion analysis, in the financial decision support domain, and to propose a general design of such knowledge enhanced FDSS.2.Introduction2.1.Research questionIn big data era, people are facing much more challenging and complicated decision making problems than before. These problems involve information from various aspects of financial market. To help financial professionals make better decisions, effective financialdecision support systems always incorporate the analysis of many kinds of market information, from market activity and movement data, company internal managerial information to latest financial news for a particular market or company.There were many researches on adopting various natural language processing techniques to the examination on market information such as financial news (Schumaker and Chen, 2009), annual reports (Li, 2006) and online discussion board (Das and Chen, 2007), in use of predicting market movement or stock price movement. One of the most popular topic is the application of text mining techniques on textual financial information to predict the performance of individual stock or the movement of the market. At a macro economy level, Peramunetilleke et al. used news headlines to predict currency exchange rates (Peramunetilleke and Wong, 2002). On a company level, Smailovic et al. examined tweets to predict stock price changes (Smailovic et al., 2012). There were also many specific application research of using test mining techniques like sentiment analysis to examine financial information. Ruiz-Martinez et al. proposed a semantic-based sentiment analysis algorithm to conduct semantic sentiment analysis on financial news, by assigning different degrees of positivity or negativity to annotated terms, to calculate the polarity of the news (Ruiz-Martinez, Valencia-García and Garcia-Sanchez, 2012). But in web environment, these partial application of natural language processing techniques on one or few sources of financial information is no longer efficient and accurate enough for financial decision making.An important characteristic of Web is the social web. Users interact with other users or applications to share their perspectives, opinions and thoughts. They are no longer a user of the application or website but also a participant of the social web. The growth of user-generatedcontent is exponential. Web content is no longer discrete but with strong interactive relationship. Although many different applications of natural language processing in financial information analysis systems yield positive results, they mainly rely on the text mining techniques to analyze financial information linguistically, calculate polarity and sentiment scores. Its limitation become more obvious in analyzing huge amount of interconnected Web content.，One of the most comprehensive methods to examine financial related information under Web environment is conducting public opinion analysis with various information source. Big data public opinion analysis is in the forefront of big data and public opinion analysis research. It use test mining and big data process related techniques to extract useful knowledge from mass data in support of decision making. It is critical for financial professions or senior management of company to gather comprehensive public opinion knowledge on related subjects as well as summarization and insights on financial market information.Nowadays, decision support modules are always embedded in public opinion analysis systems, to give decision support evidence through visual presentation. If public opinion analysis system can construct an expert knowledge base which can be further incorporated in financial decision support system to give out intellectual strategy suggestion, it would improve the efficiency and accuracy of financial decision making process. Yet there was no systematic research study on this area.Driven by the motivation to make up this research gap, my proposed research questions listed as follows:1.Whether it is feasible for big data public opinion analysis systemto construct public opinion knowledge base2.Whether it is feasible to incorporate the public opinion knowledgein the market information analysis process3.Whether the public opinion knowledge enhanced FDSSs can provide moreuseful and accurate evidence to support financial decision making 4.If the answers of the first three question are yes, then what isthe general design for the public opinion knowledge enhanced financial decision support systemReferring to the previous studies on big data public opinion analysis and decision support system (Cheng et al., 2010; 陈 and 曹, 2013; Cao et al., 2014), I believe it is feasible to construct expert public opinion knowledge base and apply it to financial decision support system with a hierarchical model. As a design science research, this study will aim to propose a general design for the public opinion knowledge base enhanced FDSS as a general system design guidance.2.2.Research methodologiesMethodology is the systematic, theoretical research strategy that comprises the rationale of the research, paradigm, theoretical model, phases and quantitative or qualitative techniques and the criteria used to interpreting data and reach the conclusion (Bailey, 1978; Irny and Rose, 2005). The paradigm of design science of information system require researchers to understand the important existing problems, explore its theoretical insights and offer solutions through building and evaluating innovative IT artifacts (Hevner, Match et al. 2004). As a design science research, this study will follow the paradigm with the final purpose of building a feasible prototype of public opinionknowledge enhanced FDSS.￥Given the questions proposed in previous section, this study will follow the framework of design science to solve them. Firstly, a thorough theoretical investigation will be conducted to examine the feasibility and effectiveness of such knowledge enhanced FDSS. From existing literatures, I will compare a variety of big data processing related techniques that has been used in public opinion analysis and evaluate their feasibility and efficiency when applied to construct knowledge base for the use of financial decision support. Related big data analysis techniques fall into following categories: data collecting techniques, internet public opinion topic detection and tracking (TDT), hotspot evaluation and analyzing and processing. Most effective and efficient big data processing techniques will be chosen, combined with behavior analysis methods, to build proposed FDSS. [Selection criteria]Secondly, dealing with the fourth question, this study will propose a general design for public opinion knowledge enhanced FDSS. The general design of such FDSS will follow the framework proposed by Walls et al. (Wall, 2004), which includes meta-requirements, meta-design and kernel theories. The meta-requirements describe what goals the theory will be applied for. It is preliminarily set as followings: the need to analyze the market information, including user-generated content, the need to incorporate public opinion knowledge and the need to effectively and efficiently support financial decision making. Meta-design describe the hypothesized system designed to meet meta-requirements. Kernel theories refer to design theories supported by justificatory knowledge from the natural or social or design science (Walls, Widemeyer, and ElSawy, 1992). The general design will include the design of conceptual model, the design of system architecture the design of reasoning model and hierarchical model integration.In the end, a prototype system will be developed based on general design to test the feasibility, effectiveness and efficiency. We will follow the methodology, and propose a general design for public opinion knowledge enhanced FDSS. To test its effectiveness and efficiency, a group of hypotheses will be given out and examined by a series of experiments. For example, we can set Hong Kong as the target market and let the system conduct public opinion analysis with information from various local or global sources on specific domain. System will generate a suggest strategy on such subject. To compare the results yielded from suggested strategy and the actual situation would prove the system is effective and efficient or not.3.Literature Review3.1.Public opinion and financial decision supportPublic opinion was first introduced back in seventeenth century by John Locke, whose work contains an early consideration of the importance of public opinion in the ordering of politics. From existing studies, we can divided social public opinion into three eras: traditional social public opinion analysis, internet public opinion analysis and big data public opinion analysis. Traditional social public opinion analysis tends to analyze the influence of hot social issues, policies and new legal provisions on social crowds (Bachner and Hill, 2014). MacLennan et al. (2012) studied New Zealand people’s attitudes to alcohol policies by the way of sampling survey. The rise of new social media platform set off the fever of internet public opinion analysis. Ceronet al. (2013) correctly predicted the president election of France in 2012 by using online public opinion data collected from Twitter, which proved that social media had a potential predictive ability.Big data public opinion analysis is one of the forefront research topic of public opinion analysis. It mainly use data mining and big data related technique to extract useful knowledge from massive amount of data for decision support (Power, 2002; Chen, Chiang and Storey, 2012). Exiting literature is more focused on big data public opinion related opportunities, challenges and research methods. There are less studies on big data public opinion from technical aspect. Broniatowski, Paul and Dredze (2014)，Li Biao (2013) and Li Xiguang (2014) narrated existing problems of big data public opinion analysis from a theoretical angle and gave detail description of public opinion information collection, study and judge and early warning functions under big data surroundings. Yu (2013) conducted public opinion analysis on China society by adopting big data analysis techniques on Baidu keywords. Merman and Scharkow (2013), Ji (2012) et al. detailed studied big data processing techniques such as parallel processing, cloud computing, advance machine learning and intellectual data processing. Li Jinhai (2014) proposed a text mining model for internet public opinion analysis using big data theory, whose accuracy and timeliness was proved through tests.As for the decision support aspect, tradition decision support modules are based on model libraries, giving visual presentation to decision makers by calling module functions. For internet public opinion analysis, there were already many commercialized systems, such as Twelvefold, Buzz Metrics, Reputation Defender. Similar Chinese products including 乐思, 云鸽, 人大方正and 优讯. There was no complete system of researching methods on big data public opinionanalysis. Most of existing studies are combining bid data characteristic related analysis techniques with methods inherited from internet public opinion analysis phrase (李, 何 and 熊, 2014; Sun, Ye and Ren, 2014).With the rapid development of internet application to financial domains, the usage of public opinion analysis is now no longer limited to pure social science field. Along with big data trend, it has now stepped into financial investment field, which are more professional and the need of accuracy, effectiveness and timeliness are much urgent more than ever. Public opinion can affect real financial market, institutions even macroeconomic through certain mechanism (Hester and Gibson, 2003). In order to protect and maintain proper order for financial market, financial regulatory agencies need to adopt financial public opinion based financial regulatory measures. Investors and financial agencies can also utilize public opinion analysis on financial market or certain hot issues about certain company or domain to predict future movement on that subject. With the development of internet and big data trend, the influence of internet financial information especially public opinion on financial market attracts more and more attention (Kiousis, Popescu and Mitrook, 2007).3.2.Decision support model for public opinion analysis in bigdata environmentThere are few studies regarding social public opinion based decision support system. In existing literature, decision support module is always located at the service layer of public opinion monitoring system, providing evidence for decision making via visual presentation (Cheng et al., 2010; Cao et al., 2014). Currently, the main functions of decision support module of public opinion monitoring system includinganalyzing public opinion information using preset models, detect and track public opinion topics and its development, automatically summarizing statistic information and decision support, etc. For existing public opinion monitoring systems, the final response strategies are always made by professional analysts, lacking intellectual strategy recommendation system. For example, the models proposed by Feng Cao (2014) and Ding and Xu (2009), which can be seen as decision support systems extracted from decision support module from public opinion monitoring system, still cannot automatically generate response strategy.）Based on studies of Feng Cao (2014), Cheng (2010) and Chen (2013), Xia and Zhen (2015) proposed a preliminary hierarchical model for public opinion analysis and decision support system (Figure 1). This modelFigure 1has three modules: data collection and storage, data processing and decision support. It added big data processing and storing technologies to existing public opinion analysis system, and most importantly, added expert knowledge base to decision support module. By adopting big data processing techniques and expert knowledge bases, this model will be able to give out recommendation strategies other than passively providing decision support evidences. Xia and Zhen’s proposed model separated decision support function from public opinion analysis system to be an individual module, and added big data technologies and expert knowledge base into the consideration. Though there is still a gap between this model and ideal design of such model applied in financial domain, this model may exerts great influence and help to this study.4.Objective and Value4.1.Research objectiveProposed detailed objectives are listed as follows:1.To conduct theoretical investigation of the feasibility of publicopinion knowledge enhanced FDSS. The scope will include but not limited to previous studies review, evaluation on technical feasibility, effectiveness and efficiency and conceptual model feasibility.2.To develop a general design for public opinion knowledge enhancedFDSS.3.To develop prototype system including a conceptual model andframework, which should be able to be embed into a domain used for specified financial decision making process.4.To demonstrate the usage of proposed prototype system and itsconceptual model and framework in specific domains. Proposed model and framework need to be validated and evaluated in real-time test.Abstract model design cannot guarantee the effectiveness and efficiency of such system in a particular domain.4.2.Research valueAs mentioned in previous section, public opinion analysis has stepped out the limitation of social science field. The application of it to a more professional financial domain is attracting more and more attention. Three possible and most useful applications of big data public opinion analysis in financial domain are listed as follows:1.…2.Investment supportIt is not surprising that many people already see public opinion as a new tool helping them invest. Some new event or a heated discussion happened somewhere on internet may have heavy influence on investors. It affects their investing behavior, confidence on certain market or company. A financial decision support system with big data public opinion analysis technique can monitor and track financial news, relevant comments or online financial forum discussions, analyze the relationship of these information and specific financial domain and extract and organize useful knowledge from these information to provide useful and predictive evidence to support investors and agencies.3.Regulatory decision supportBig data public opinion knowledge enhanced FDSS can act as decision support tools for financial regulatory institutions, supporting traditional financial analyzing techniques byproviding online information analysis to construct comprehensive hierarchical knowledge base to support decision making on specific financial policy domain.4.Reputation and risk managementInternet can amplify public opinions on social or financial events in an irresistible way. It is crucial for companies or financial regulatory intuitions to monitor public opinions on related subjects to manage negative risk events outbreak which may damage their reputation or business. A big data public opinion knowledge enhanced FDSS can provide users with comprehensive, accurate and timely analysis of public opinion information on specific domain.ReferencesBachner, J. and Hill, K. (2014). Advances in Public Opinion and Policy Attitudes Research. Policy Stud J, 42, .Bailey, K. (1978). Methods of social research. New York: Free Press.Broniatowski, D., Paul, M. and Dredze, m. (2014). Twitter: big data opportunities. Inform, 49, .'Cao, F., zhan, Z., Jing, Y. and Guan, X. (2014). A Model of Ecological Monitoring and Response System for Internet PublicOpinion. International Journal of Multimedia and UbiquitousEngineering, 9(5), .Ceron, A., Curini, L., Iacus, S. and Porro, G. (2013). Every tweet counts How sentiment analysis of social media can improve ourknowledge of citizens' political preferences with an application to Italy and France. New Media & Society, 16(2), .Chen, H., Chiang, R. and Storey, V. (2012). Business Intelligence and Analytics: From Big Data to Big Impact. MIS quarterly, 36(4), .Cheng, X., Zhu, L., Zhu, Q. and Wang, J. (2010). The Framework of Network Public Opinion Monitoring and Analyzing System Based on Semantic Content Identification. JCIT, 5(10), .Das, S. and Chen, M. (2007). Yahoo! for Amazon: Sentiment Extraction from Small Talk on the Science, 53(9), .Ding, J. and Xu, J. (2009). IPOMS: an Internet Public Opinion Monitoring System. In: Applications of Digital Information and Web Technologies, 2009. ICADIWT '09. Second InternationalConference on the. IEEE. .Hester, J. and Gibson, R. (2003). The Economy and Second-Level Agenda Setting: A Time-Series Analysis of Economic News and PublicOpinion about the Economy. Journalism & Mass CommunicationQuarterly, 80(1), .Hevner, A., March, S., Park, J. and Ram, S. (2004). Design Science in Information Systems Quaterly, 28(1), .Irny, S. and Rose, A. (2005). Designing a Strategic Information Systems Planning Methodology for Malaysian Institutes of Higher Learning. Issues in Information System, VI(1).JI, C., LI, Y., QIU, W., JIN, Y., XU, Y., AWADA, U., LI, K. and QU, W. (2012). BIG DATA PROCESSING: BIG CHALLENGES ANDOPPORTUNITIES. J. Inter. Net., 13(03n04), .Kiousis, S., Popescu, C. and Mitrook, M. (2007). Understanding Influence on Corporate Reputation: An Examination of PublicRelations Efforts, Media Coverage, Public Opinion, and Financial Performance From an Agenda-Building and Agenda-SettingPerspective. Journal of Public Relations Research, 19(2), .—Li, F. (2006). Do Stock Market Investors Understand the Risk Sentiment of Corporate Annual Reports.SSRN Journal.Maclennan, B., Kypri, K., Langley, J. and Room, R. (2012). Public sentiment towards alcohol and local government alcohol policies in New Zealand. International Journal of Drug Policy, 23(1), .Mahrt, M. and Scharkow, M. (2013). The Value of Big Data in Digital Media Research. Journal of Broadcasting & Electronic Media,57(1), .Peramunetilleke, D. and Wong, R. (2002). Currency Exchange Rate Forecasting from News Headlines. In: Australasian DatabaseConference. Melbourne.Power, D. (2002). Decision support systems. Westport, Conn.: Quorum Books.Ruiz-Martínez, J., Valencia-García, R. and García-Sánchez, F.(2012). Semantic-Based Sentiment analysis in financialnews. Proceedings of the 1st International Workshop on Finance and Economics on the Semantic Web, .Schumaker, R. and Chen, H. (2009). Textual analysis of stock market prediction using breaking financial news. TOIS, 27(2), .Smailovic, J., Grcar, M., Znidarsic, M. and Lavrac, N. (2012).Sentiment analysis on tweets in a financial domain. In: 4thJožef Stefan International Postgraduate School StudentsConference.Sun, X., Ye, J. and Ren, F. (2014). Intelligent computing methodologies. Cham: Springer.Walls, J., Widmeyer, G. and El Sawy, O. (1992). Building an Information System design theory for vigilant ExecutiveInformation Systems. Information Systems Research, 3(1), .喻, 国明. (2013). 构建社会舆情总体判断的大数据方法———以百度海量搜索数据的处理为例. 新闻与写作, 7, .夏, 火松. and 甄, 化春. (2015). 大数据环境下舆情分析与决策支持研究文献综述. Journal of Intelligence, 34(2), , 21.李, 希光. (2014). 大数据时代的舆情研判和舆论引导. 中国广播, 5, .李, 彪. (2013). 大数据视域下社会舆情研究的新境界. 编辑之友, 6, .李, 金海., 何, 有世. and 熊, 强. (2014). 基于大数据技术的网络舆情文本挖掘研究. 情报杂志, 33(10), .陈, 彦舟. and 曹, 金璇. (2013). 基于 Hadoop 的微博舆情监控系统. 计算机系统应用, (4), .。
Does IFRS mandatory adoption influence IPO underpricing? July, 2013ABSTRACTThis study examines the impact of mandatory IFRS adoption on IPO underpricing. The latter is usually associated with information asymmetry among investors. We expected that IPO firms’ use of IFRS rather than domestic GAAP may affect IPO underpricing through two mechanisms: (1) the quality, and (2) the comparability of financial reports. We find that, mandatory IFRS adoption is not associated with a decrease in IPO underpricing and the only factor that is associated with IPO underpricing is the trend of the financial market. This analysis investigates the effects of the mandatory adoption of IAS/IFRS in an original way. It is the first Italian research that studies the effects of IAS/IFRS on the financial statements by analyzing their effects on the underpricing of IPOs. It shows that IFRS mandatory adoption does not reduce the asymmetry information among investors thus decreasing the underpricing.1.IntroductionFrom 2005 almost all publicly listed companies in Europe are required to prepare their financial statements according to International Financial Reporting Standards (IFRS).The mandatory adoption of IFRS, instead of local GAAP, in the European Union, has been the most significant challenge regarding the financial reporting. Mandatory has been the adoption of a new set of principles for the preparation of the consolidated financial statements of publicly listed companies; while discretionary has been the adoption for the individual financial statements of publicly listed companies and for those companies not publicly listed.In Italy, in 2005, the government issued the Legislative Decree No. 38/2005. It prescribes that all listed companies and companies operating in the financial sector such as banks, supervised financial companies and companies issuing financial instruments widely distributed among the public, have to apply IFRS in their consolidated financial statements from 2005 onwards and in their individual financial statements from 2006 or, optionally, from 2005. Furthermore, the application of the new standards is also allowed to other non-listed companies; they can voluntary adopt IFRS for both individual and consolidated financial statements. Only companies presenting abbreviated financial statements are excluded. As for the individual account, IFRS have been authorized to give homogeneity to the financial statements (OIC, 2005)Proponents of mandatory IFRS adoptio n state that IFRS “will enable investors to compare the financial results of companies operating in different jurisdictions more easily and provide more opportunity for investments and diversification” (Tweedie, 2006).According to Ball (2006), IFRS could facilitate cross-border comparability and increase reporting transparency, enabling stakeholders to understand the financial results of firms in the whole world. Moreover, IFRS adoption could decrease information costs. Choi and Meek (2005), state that companies might also benefit by reducing information asymmetry, giving to all investors the chance to make more efficient investment decisions, and thus lowering the cost of capital. Investors may also benefit as it could lead to more-informed valuation of equity markets reducing the risk of adverse selection for the less-informed investors.All the above benefits rely on the presumption that mandatory IFRS adoption provides superior information to market participants compared to local GAAP.So in the past years empirical research has focused the attention on the transition process from local GAAP to IFRS and on the effects of IFRS adoption on the quality and on the comparability of the financial statements (Barth et al., 2008; Lang et al., 2012; De Fond at al., 2011).The aim of this research is to verify if, referring to the IPOs case, it’s possible to have information about the over and over stated higher quality and comparability of IFRS compliant financial statements1.As is known, with the term Initial Public Offering we mean the offer to subscribe and/or sell2 securities of a company that for the first time gain admittance to the market negotiation. A critical step in the process of public placement is the fixing of the offer price of the single share and of particular interest it’s the relationship between the latter and the economic value of the share itself.Studies on IPOs in every country of the world underscore a recurring anomaly. Research, in fact, observes, for many IPOs, a phenomenon commonly called underpricing3. According to the most qualified definition of the undepricing, it is the positive spread between the first trading day closing price of the newly issued share and the offer price fixed by the underwriter4. The economic literature has tried to find a justification for this phenomena building models which explain why the offer price of the shares is set below the real value of the shares themselves. Furthermore, these studies have verified if this anomaly is found for IPOs in different countries.Many scholars ascribe IPO underpricing to information asymmetry among the actors involved in the process of listing (Rock, 1986; Beatty e Ritter, 1986; Allen and Faulhaber, 1988). Rock (1986), Beatty and Ritter (1986) attribute the underpricing to asymmetry between informed and uninformed investor; the former having invested resources to acquire information about an IPO firm’s value. Uninformed investors are at an information disadvantage and can therefore be exploited. The offer price must be fix low enough to leave “money on the table” in order to attract uninformed investor to participate in the IPO market and to compensate informed investor for the cost of acquiring information. Allen and Faulhaber (1988) assume that firms have information about the quality of their investment projects not available to external investors. Firms with the best economic perspectives try to signal their quality with a low offer price and with the number of shares they hold.Among studies that have tried to find an explanation to underpricing, few of them focused their attention on the effects that the information contained in the financial statements could have on the evaluation of the shares which will be placed on the market.Therefore, the aim of our study is to analyze the consequences that the mandatory adoption of IFRS could have had on the evaluation of those firms that decided to go public on the Italian stock market.1For details see: Armstrong C., Barth M., Jagolinzer A., and Riedl E., 2010; Market reaction to the adoption of IFRS in Europe; The Accounting Review 85, 31-61.2With the offer to sell, previously issued shares are offered; with the offer to subscribe, newly issued shares are offered; with the offer to sell and subscribe, both newly and previously issued shares are offered. When newly issued shares are offered, the operation allows to raise capitals and to increase investors.3 Following prior studies, we measure the degree of IPO underpricing by the first trading day stock return relative to the offer price.4The underwriter is the investment bank, involved in the placement consortium, which handles, as supervisor, with the share placement. Usually the underwriter is, at the same time Bookrunner, sponsor and Global Coordinator of the IPO.The interest of this analysis lies in the fact that, since the shares have never been listed before, the only information available to investors are those contained in the prospectus; thus it is possible to analyze the effects that IFRS adoption has on the evaluation on the IPO company.Our study makes some important contributions. First, it complements the growing body of literature that examines economic consequences of mandatory IFRS adoption.Most studies that analyze the effects of the mandatory adoption of IFRS, focus on setting such as analysts’ infor mation environment and foreign mutual fund investment (e.g. Yu, 2010; Byard et al., 2010; DeFond et al., 2011; Tan et al.,2011). These studies, however, are silent on whether mandatory IFRS adoption influences the equity issuance process. An exception is Wang and Welker (2011) who investigate the timing of equity issuances during the transition period to IFRS. Our study complements Wang and Welker (2011) by focusing on the impact of IFRS mandatory adoption on IPO underpricing.Second, we extend the extent literature on underpricing by documenting the effect of changes in accounting standards on the initial return. Consistent with the information asymmetry explanation of underpricing, prior study find lower underpricing with greater disclosure (Leone et al., 2007; Bulton et al., 2011).We expect IFRS to affect IPO underpricing through two mechanisms: reporting quality and the comparability of financial reports. First, as global set of accounting standards, supposedly provide improved reporting quality compared to local GAAPs on average, which in turn may reduce IPO underpricing5. Second, IFRS adoption may reduce IPO underpricing by increasing the comparability of financial reports among industry peers that use the same accounting standards. Comparability of fi nancial reports is defined as “the quality of information that enables users to identify similarities and differences between two sets of economic phenomena” (FASB, 2008). Enhanced comparability may facilitate investors and underwriters in making comparati ve assessments of firms’ accounting performance and share prices. That is, enhanced comparability may allow investors to clarify similarities and differences among industry peers and help investors to extract useful information, which would decrease information asymmetry. As a result, enhanced comparability may mitigate IPO underpricing. The sample used in our study is composed by all the companies that went public from 2000 to 2012 excluding companies whose data were not available. The results of our study show that the adoption of IFRS does not decrease IPO underpricing. Therefore, it seems that the introduction of a new set of principles, such as IFRS, does not improve the quality and comparability of the financial statements. This conclusion is in line with previous research that underscored how the mere introduction of a new set of standards is not sufficient to change the quality of accounting. The rest of the paper is organized as follows. Section 2 reviews previous literature and develops hypothesis. Section 3 discusses the research design and empirical results. Section 4 concludes.2.Prior studies5While many countries have adopted IFRS with the expectation that IFRS would benefit financial statement users (Barth, Landsman, and Lang, 2008), it is unclear whether the use of IFRS actually provides improved reporting quality compared with domestic GAAPs (Ahmed, Neel, and Wang, 2010; Capkun, Collins, and Jeanjean, 2011).2.1IPO underpricingThe Initial Public Offerings have been widely studied by the literature. By far the most relevant finding has been the phenomenon of underpricing. Not only does the literature document underpricing since 1970, it also provides evidence of underpricing in countries around the globe (Ibbotson, 1975; Loughran et al., 1994). A lot of scholars wonder why a company decides to “leave money on the table” going public at a price fixed below the real value of the company itself.Theoretical studies attribute IPO underpricing to information asymmetry among the actors of the process of listing about the value of the IPO firm6.Rock (1986) shows that when some investors are better informed than others, underpricing becomes necessary to convince uninformed investors to buy shares of the IPO firm. In particular, when shares are overpriced, uninformed investors receive a full allocation of shares because informed investors withdraw from the market. In contrast, when the offering is underpriced, uninformed investors receive only an allocation of rationed shares because informed investors stay in the market. Since uninformed investors do not participate in the offering until the price falls enough to compensate for this adverse selection cost, firms have to issue their share at a discount underpricing their offering.Ritter (1984) refers to R ock’s model (1982). He states that firms, underwriters and investors are not sure about the true value of the firm. Investors, incurring costs, can know the true value of the shares; such investors are called informed investors. Ritter affirms that underpricing is the remuneration, for informed investors, for incurring costs to become informed; therefore higher is the uncertainty about the value of the firm, higher is the remuneration necessary to compensate uninformed investors for the costs incurred.Betty and Ritter (1986), using Rock’s model (1982), show the existence of a relationship between underpricing and ex-ante uncertainty about the true value of the firm. They state that higher is the uncertainty about the true value of the firm, higher is the risk born by uninformed investors for investing in the firm. In order to maintain uninformed investors in the market, it’s necessary compensating them with a high expected return by fixing for shares an offer price below their real value.Allen and Faulhaber (1988), assume that firms are better informed than external investors because they know the quality of their investment project. Firms with good future prospects 6There are other theories that explain underpricing referring to the relationship between issuer and underwriter. Baron and Holmstrom (1980) underline the conflict of interest between issuer and underwriter. They state that the underwriter is motivated to fix a low offer price to limit his commitment and the costs in the marketing phase, while the issuer wants the price to be high as much as possible to maximize the profit. Boron (1982), in his model, describes the underwriter as he who has information about the potential demand and about market condition that are not available to the issuer. Furthermore, he, such as Baron and Holmstrom (1980) characterizes the relationship between issuer and underwriter as an agency relationship where the issuer cannot directly control the underwriter in the marketing and distribution phase. Finally, Shiller (1990) affirms that IPO underpricing is generated by the bank that takes care of the placement to create the pretense of an over demand.signal their higher quality with a lower offer price and with the quantity of shares that they hold. Fixing an offer price below the real value, can be a believable signal of the quality of the firm because only “good” firms can recover the cost of underpricing with further placement.2.2IFRS and IPO underpricingIPO firms’ use of IFRS rather than domestic GAAP may affect IPO underpricing through two mechanisms: reporting quality, and the comparability of financial reports. First, IFRS may reduce IPO underpricing if IFRS provide higher reporting quality than domestic GAAP. Barth et al. define accounting quality as less earnings management, more timely loss recognition and a higher association of accounting amounts with share prices. The literature gives different definition of accounting quality (for an analysis, see Prencipe, 2006).However, academic research comparing reporting quality under IFRS with that under domestic GAAPs provides mixed results (Barth et al., 2008; Ahmed et al., 2010). Barth et al. (2008) find that firms applying IFRS from 21 countries generally financial reports of higher quality.Daske and Gebhardt (2006), analyzing a sample of Austrian, German and Swiss firms, find an improvement of the disclosure after IFRS adoption.Tsalavoutas and Evans (2010) observe that in Greece, the transition to IFRS has limited occasion for creative accounting.Differently from the studies above mentioned, there are others that find no relationship or a negative relationship between the adoption of IFRS and the quality of the financial statements. Barth at al. (2008) although find that IFRS adoption was associated with higher accounting quality, states that this may not be always true. First IFRS could be of lower quality than domestic GAAP and the flexibility of principle-based standards, such as IFRS, could provide greater opportunity for manager to manipulate earnings. Second, there are other factors that could influence the effect of IFRS adoption.Burgstahler, Hail and Leuz (2006), Ball, Robin and Wu (2003) suggest that lax enforcement can result in limited compliance with the standards, thus influencing their effectiveness.Ahmed et al. (2010) find that accounting quality deteriorates following mandatory IFRS adoption in 2005. They argue that principles-based IFRS increase opportunities for managers to exercise discretion rather than faithfully report underlying firm value. They show that firms mandated to adopt IFRS increase income smoothing and aggressive accruals and decrease timely loss recognition.Vantendeloo and Vanstralen (2005), studying a sample of German firms that voluntary adopted IFRS, do not find significant difference between their financial statements and those of firms that used domestic GAAP in terms of accounting quality.Cameran and Campa (2012) analyze 301 italian not listed companies that adopt IFRS. They analize the effect of IFRS mandatory adoption on the earnings quality. Their results do not show any improvement in earnings quality following the introduction of IFRS.The second mechanism through which IFRS adoption could decreases IPO underpricing is by enhancing the comparability of the financial statements. Thanks to an increased comparability, investors can compare firms located all over the world. The chance to compare similarities and differences between firms operating in the same industry all over the world should reduce asymmetry information among investors. However, there are some reasons why this may not happen. First, if the reporting quality of firms using IFRS is lower than that of firms using domestic GAAP (Ahmed et al., 2010), then comparability under IFRS could be lower than that under domestic GAAP. Second, it might be difficult for investors to compare IFRS-reporting firms domiciled in different jurisdictions. Different jurisdictions provide heterogeneous levels of investor protection and reporting incentives for firms (Leuz, Nanda, and Wysocki, 2003). In contrast, firms using a set of domestic GAAP by definition consist of relatively homogenous firms in the same jurisdiction. Third, comparability among firms using IFRS will be further limited if there are variations in country-level and firm-level interpretations of IFRS.Results of analysis about the effects on the financial statements’ comparability of the IFRS mandatory adoption are mixed too.De Fond et al. (2011), find an improvement of the comparability after IFRS mandatory adoption. The improvement of the comparability is measured by the increase of international investments.Lang et al. (2011), on the other hand, underscore an unchanged level of comparability after IFRS mandatory adoption.The effects on the quality and the comparability of IFRS mandatory adoption are influenced by the quality of the enforcement that is the emanation and the application of laws that guide IFRS application. The latter have to limit managers’ discretionary and opportunism (Leuz, Nanda and Wysocki, 2003).According with what stated above, several studies show that IFRS adoption increases the quality and the comparability of the financial statements only in countries where the enforcement is of high quality (Daske et al., 2008; Byard et al., 2011).Although IFRS may improve firms’ reporting quality compared with domestic GAAP, this will happen only if firms actually comply with the requirement of IFRS.About the comparability, the enforcement enviroments in which comparable firms are listed are also critical to the comparability between an IPO firms and its comparable firms. Comparability under IFRS is likely to be realized if the comparable firms listed in other jurisdiction are faithfully following IFRS.3.Hypothesis development and study designOur hypothesis is about the validity for the market of the information provided by the financial statements in order to evaluate the shares. If the information is characterized by high quality, the evaluation of the shares given by the market should not differ from the fixed price arising from information of the financial statements. So, referring to Rock’s model (1986), investors shouldnot incur costs to obtain information about the firm, thus becoming “informed” investors, given that information are public and available to all investors.Based on the above reasoning, we hypothesize the following:Underpricing of shares of IPO firms adopting IFRS prior to the listing is lower than IPO underpricing of IPO firms adopting Italian GAAP prior to the listing.The database used in this study is composed by all the firms that decided to go public on the Italian market from 2000 to 2012. In this period most firms used book-building to go public; this mechanism allows the price to be “adjusted” on the base of the de mand. From 1994 Italian firms use this mechanism to fix the offer price (Paleari, Cassia and Redondi, 2004)7. With the book-building the offer price varies in a fixed range indicated in the prospectus, but cases in which the price is fixed outside this range are not rare. During bookbuilding, underwriters collect orders in terms quantity and price or without price limit in an electronic book. In this way they create a demand curve composed by all the orders and the prices. According to the curve a price will be negotiated between the IPO firm and the underwriter.Sometimes, IPOs consist of the offer of shares already owned by existing shareholders while, in other cases, shares are issued in conjunction with the IPO: in the latter case with the IPO new capital is collected.We excluded from the sample those firms whose data were not available. The sample is composed by 141 firms: 71 adopting Italian GAAP to prepare their financial statements prior to the listing and 70 adopting IFRS to prepare their financial statements prior to the listing.The period length arise from the necessity to make the sample as numerous as possible. Firms that went public before 2000 have not been considering due to the unavailability of the data.The database has been built using information available in the prospectus prepared for the IPO and on the Borsa Italiana website.Tab. 1 Main characteristics of the Italian IPO from 2000 to 20127 Three mechanisms can be used to place shares: the auction, the fixed price offer and the book-building. In fixed price offers, the price and allocation rules are set before demand is received, and shares are allocated according to the rules announced earlier.With book building, the underwriter typically arranges for investors to attend a road show and then collects indications of interest, which are used to build the order book. The offering price is set only after the order book is full, giving the underwriter some idea of demand.Auctions for IPOs have taken several forms. Uniform price auctions are multi-unit sealed bid auctions in which all winning bidders pay the same price. The price paid may be the market-clearing price (the highest price that allows all shares to be sold), or it may be below th e clearing price. A ”dirty” IPO auction is a uniform price auction where they ”leave something on the table” by pr icing below market-clearing. In a discriminatory or pay-what-you-bid auction, each winning bidder pays his or her own bid.The name of the companies and the distribution of the information per years are contained in the Tab. 1. Furthermore, in this table are contained the main characteristics of the firms that decided to go public from 2000 to 2012.From the distribution we can observe that the most fruitful year has been 2000 with 42 IPOs. From 2008 we find a decline in the number of IPOs probably due to the financial crisis that from 2008 has affected the whole world as a consequence of the crisis of 2007 in the United States.Based on the data released by the European Observatory on IPO, the first nine months have been characterized a drop in the number and the value of IPOs. In particular, 271 IPOs have been realized. 53% less compared to the first nine month of 2007. This drop represents the most significant since Observatory’s launch in 2002 and confirm the strong relationship between the trend of the financial market and the decision to go public.The difference between the offer price and the closing price on the first day of trading (initial return) and the variation of the market index on the same day (market return) are expressed in percentage.The firms that went public adopting Italian GAAP show an underpricing on average of 5,9%, while firms that decided to go public adopting IFRS show an underpricing on average of 7,9%. These levels of underpricing are lower than those observed by Paleari, Cassa Redondi (2004) that find an underpricing of 15,46% on average from 1991 to 2001.The underpricing in Italy is, on average, lower than US underpricing based on the date provided by Ritter (2008) that find an underpricing of 22,3% on average from 1990 to 2008.Firms that decided to go public from adopting Italian GAAP show an average age of 27 years. However, there is and high variability with extreme values (143 and 0) that could alter the mean. The firms that decide to go public adopting IFRS show an average age of 20 years. In this case too there is a high variability with maximum of 172 and minimum of 0 for firms that went public just after the constitution.2.3Descriptive statisticThe Underpricing, the main variable of interest, is represented by the difference between the offer price and the closing price on the first day of trading, on the secondary market, expressed in percentage (Bulton ae al., 2011). The market return is the variation of the market index on the same day. The age has been determined as the years between the constitution and the listing. With offer price we intend the product between the number of shares and their offer price.Tab. 2 Decriptive statisticInitial Return-------------------------------------------------------------Percentiles Smallest1% -.1259048 -.14243325% -.0911765 -.125904810% -.0581081 -.1238889 Obs 14225% -.011 -.1183333 Sum of Wgt. 142 50% .0299545 Mean .0692942Largest Std. Dev. .140314275% .1273913 .493111190% .24625 .4967742 Variance .0196881 95% .3666667 .5925926 Skewness 1.58729 99% .5925926 .6334286 Kurtosis 5.941288Market Return-------------------------------------------------------------Percentiles Smallest1% -.0250944 -.2821075% -.0161822 -.025094410% -.0130731 -.0236842 Obs 142 25% -.0068428 -.0194738 Sum of Wgt. 142 50% .0002971 Mean -.0024976Largest Std. Dev. .025355375% .0041534 .016150990% .0102292 .016521 Variance .0006429 95% .014235 .0284405 Skewness -9.501597 99% .0284405 .0354987 Kurtosis 105.7156Age-------------------------------------------------------------Percentiles Smallest1% 0 05% 1 010% 1 0 Obs 14225% 5 1 Sum of Wgt. 14250% 16 Mean 23.23239Largest Std. Dev. 27.7888975% 29 9990% 55 104 Variance 772.2222 95% 76 143 Skewness 2.46967 99% 143 172 Kurtosis 10.62947Offer Size-------------------------------------------------------------Percentiles Smallest1% 296250 1000005% 7875000 29625010% 1.53e+07 3380000 Obs 141 25% 2.94e+07 3600000 Sum of Wgt. 141 50% 7.74e+07 Mean 1.69e+08Largest Std. Dev. 3.41e+0875% 1.54e+08 1.52e+0990% 3.33e+08 1.92e+09 Variance 1.16e+1795% 6.14e+08 2.07e+09 Skewness 4.49603699% 2.07e+09 2.26e+09 Kurtosis 24.47814The Tab. 2 shows the descriptive statistics of the variables used in the regression model. The Tab. 3 shows the T-Test to compare the average underpricing for IFRS adopters and the average underpricing for Italian GAAP adopters.Tab. 3 Two sample T-Test with unequal variancesdiff = mean(0) - mean(1) t = -0.9773Ha: diff < 0 Ha: diff != 0 Ha: diff > 0Pr(T < t) = 0.1651 Pr(|T| > |t|) = 0.3301 Pr(T > t) = 0.8349The T-Test shows that the difference between the means is not significant. This suggests that the IFRS mandatory adoption could not have improved the quality and the comparability of the financial statements.2.4Regression modelWe use a multiple regression model to verify if IFRS mandatory adoption decreases underpricing. We use underpricing as dependent variable and age, initial return, offer size, IFRS adoption as independent variable. For IFRS we use a dummy variable equal to 0 for Italian GAAP adopters and 1 otherwise. Results show that the only variable that influence undepricing is the market return.Tab. 4 Multiple regression modelNumber of obs = 141Prob > F = 0.1500R-squared = 0.0480Root MSE = .13826。
