< PreviousA PROFESSIONAL DEVELOPMENT JOURNAL for the CONSULTING DISCIPLINES10 JANUARY | FEBRUARY 2019 t h e v a l u e e x a m i n e rAs shown on Table 3, the largest the largest ICO on record, EOS, raised nearly $4.2 billion. EOS is a platform that aims to make it easier for developers to make decentralized applications. The top industries that experienced ICOs in 2018 were infrastructure, finance, and communications. While infrastructure was also the top industry for ICOs in 2016 and 2017, trading and investing ICOs that were popular in 2016 became less popular by 2018.While the significant fluctuation in the price of Bitcoin has been a change in the overall exuberance investors had in the concept vs. any perception of value, the value of an ICO lies in the perceived supply and demand of the underlying technology and ecosystem upon which the ICO is based. Fundamental value increases in tokens issued in an ICO are based on the maturation of the issuer’s platform, which increases the demand for the tokens. The initial value of tokens offered in an ICO is more about the marketing efforts of the issuer’s whitepaper, the participation by known professional investors, and a belief in the ecosystem. TRENDS IN 2019Undoubtedly, 2018 was a challenging year for the crypto economy. While all digital currencies experienced a roller-coaster ride in terms of their prices, Bitcoin saw the most significant fluctuations. However, according to experts quoted in are exaggerated. Here are some trends they see: STOs Will Replace ICOsRumor has it that ICOs are dead. Writing in Forbes,4 Charles Bovaird cites several crypto market experts pronouncing the death sentence and introducing the next big thing to inherit ICOs mantle: Security Token Offerings (STO). STOs have the potential to make securities on blockchain more transferable and transparent. STOs minimize the chances of fraud and provide investors with some protection.Custody Solutions Will Get CheaperCustody solutions are third-party providers of security and storage services for digital currencies. They are turning out to be the industry standard when it comes to cryptocurrency lending. As they have clear regulations and provide a model way to protect crypto assets, investors are showing great interest.4 Charles Bovaird, Is The ICO Market Truly Dead? Forbes, January 16, 2019, forbes.comTABLE 3: TOTAL RAISED FROM ICOSA PROFESSIONAL DEVELOPMENT JOURNAL for the CONSULTING DISCIPLINESt h e v a l u e e x a m i n e r JANUARY | FEBRUARY 2019 11Cryptocurrencies Will Become MainstreamAs Bitcoin futures gains mainstream adoption, investors are becoming more confident in betting against them and settling the contracts in regular currency. Moreover, investors can trade Bitcoins without actually owning them.Similarly, as all cryptocurrencies become mainstream, there will be a considerable increase in the number of crypto-backed loans. These loans will be an alternative to conventional loans, and people will be able to get a crypto-backed loan at a low cost, without any credit checks.Cryptocurrency Trading Volume Will Grow Fifty PercentWhile uncertainty still mainly circles the cryptocurrency sector, its overall trading volume is projected to increase in 2019 significantly, a new report by the Satis Group5 shows. 5 Crypto Asset Market Coverage Initiation: Trading & Custody, September 18, 2018, d3h2iTlKWIa4FTLKGJsUn3mis5gThe study, Crypto Asset Market Coverage Initiation: Trading & Custody, highlights that cryptocurrency trading volume will climb fifty percent next year and will continue to grow at a compound annual growth rate (CAGR) of nine percent until 2028.Government Agencies Will Consider Crypto PaymentsBitcoins have been recognized as a mode of payment for goods and services by merchants and some e-commerce websites across the globe for quite a few years. After merchants, government agencies are also embracing the concept of using cryptocurrency as a mode to pay for the purchases.After the State of Ohio allowed its citizens to pay their taxes in Bitcoin, the U.K. indicated it was open to its residents paying their bills and taxes using Bitcoin. It is likely other governments will follow suit. The Value Examiner®—May/June 2016 CPE ExamOffice Use Only: Invoice #: Examiner CPE Rev 7/14/16–Page 1Earn five hours of NACVA CPE*by reading The Value Examiner and For CPE credit, scan and e-mail to: (801) 486-7500,or mail to: 5217 South , UT 84107Member cost: $76.50 (Non-Member cost: $85.00) Name: Designations: NACVA Member #: Firm Name: IBA Member #:Address: City: State: ZIP: Tel: Fax: E-mail: Check #: (payable to: NACVA)or VISA MasterCard AMEX Discover Diners ClubCredit/Debit Card #: Expiration Date: Credit card billing address: Same, orAddress: City: State: ZIP:Authorized Signature‡Date: ‡By signing, you authorize the National Association of Certified Valuators and Analysts (NACVA) to charge your account for the amount indicated. NACVA can also initiate in the event a credit or correction is due. Your signature authorizes NACVA to confirm the use either for future communication. NACVA will not disclose or share this information * This exam does not qualify for NASBA QAS CPE credit.Important note: Although this exam qualifies for NACVA CPE, it may not be accepted by all state boards or accrediting organizations. Therefore, individuals should contact their state board or accrediting to determine if passing an exam after reading a book/magazine meets their CPE State CPE Sponsor #:_______________.Does the IPCPL Make Sense (Part II) By Richard R. Conn, CMA, MBA, CPA, ABV, ERP1. The IPCPL methodology is founded upon the premise that there is a direct inverse relationship between the firm size (i.e., Enterprise Value) and cost of capital—the smaller the firm, the higher its risk rate. In Part II of his continuing argument against IPCPL, Conn takes the position that:a.He agrees with the premiseb. He disagrees with the premisec.He offers no comments either in support of or against the concept2. BB&D and Gorshunov are really saying both that smaller EV firms have higher costs of capital and that there is a direct correlation between firm EV and its revenues (e.g., smaller firms have lower revenues). However, Conn’s regressions of the actual IPCPL data has led him to conclude that:a.There is a very strong inverse correlation between firm EV and its cost of capital (i.e., smaller EV firms have higher risk rates)b. There is a very strong direct correlation between firm revenues and EV size (i.e., firms with higher revenues have greater EV’s than firms with lower revenues)c.There is no correlation between firm EV and its cost of capital and, at best, only very weak correlation between firm revenues and EV size d. High degrees of correlation is not necessarily an indication of causalityThe Value Examiner®—March/April 2016 CPE ExamOffice Use Only: Invoice #: Examiner CPE Rev 5/4/16–Page 1Earn five hours of NACVA CPE*by reading The and completing this exam.For CPE credit, scan and e-mail to: (801) 486-7500,or mail to: 5217 South State Street, Suite 400, 84107Member cost: $76.50 (Non-Member cost: $85.00) Name: Designations: NACVA Member #: Firm Name: IBA Member #:Address: City: State: ZIP: Tel: Fax: E-mail: Check #: (payable to: NACVA)or VISA MasterCard AMEX Discover Diners ClubCredit/Debit Card #: Expiration Date: Credit card billing address: Same, orAddress: City: State: ZIP:Authorized Signature‡Date: ‡By signing, you authorize the National Association of Certified Valuators and Analysts (NACVA) to charge your account for the amount indicated. NACVA can also initiate credit entries to your account in the event a credit or correction is due. Your signature authorizes NACVA to confirm the above information via e-mail and/or fax and to use either for future communication. NACVA will not disclose or share this information with third parties.* This exam does not qualify for NASBA QAS CPE credit.Important note: Although this exam qualifies for NACVA CPE, it may not be accepted by all state boards or accrediting organizations. Therefore, individuals should contact their state board or accrediting organization to determine if passing an exam after reading a book/magazine meets their CPE requirements. State CPE Sponsor #:_______________.A New Era for Fair Market Value Physician CompensationBy Mark O. Dietrich, CPA, ABV1. Appraisal practice and government enforcement surveys have been employed as a “gold standard” in measuring fair market value for physician compensation for many years. In this article, the author makes the case that this measurement:a. Is the most efficient and accurate way to measure FMV for physician compensationb. Is based on a series of critically flawed beliefs amongst many regulators and appraisersc. Is flawed, but still usefuld. None of the above2. Regarding the question whether or not all physicians will soon be employed by hospitals, the author suggests: to single specialty physicians in private practice considering are the chief employer of specialty physicians in private practicec. Survey data does not exist to support either conclusiond. More research needs to be done to establish specialty physicians considering private practiceThe Value Examiner®—July/August 2016 CPE ExamOffice Use Only: Invoice #: Examiner CPE Rev 8/31/16–Page 1Earn five hours of NACVA CPE*by reading The Value Examiner and completing this exam.For CPE credit, scan and e-mail to: fax to: (801) 486-7500;or mail to: 5217 South State Street, Suite 400, Salt Lake City, UT 84107Member cost: $76.50 (Non-Member cost: $85.00)Announcing—The Value Examiner CPE exam can now be taken online! Visit the exam. There, you will be able to purchase, complete, and earn five hours of NACVA CPE*. You will instantly receive a certificate of completion for each exam you pass. Name: Designations: NACVA Member #: Firm Name: IBA Member #:Address: City: State: ZIP: Tel: Fax: E-mail: Check #: (payable to: NACVA)or VISA MasterCard AMEX DiscoverCredit/Debit Card #: Expiration Date: Credit card billing address: Same, orAddress: City: State: ZIP:Authorized Signature‡Date: ‡By signing, you authorize the National Association of Certified Valuators and Analysts (NACVA) to charge your account for the amount indicated. NACVA can also initiate credit entries to your account in the event a credit or correction is due. Your signature authorizes NACVA to confirm the above information via e-mail and/or fax and to use either for future communication. NACVA will not disclose or share this information with third parties.* This exam does not qualify for NASBA QAS CPE credit.Important note: Although this exam qualifies for NACVA CPE, it may not be accepted by all state boards or accrediting organizations. Therefore, individuals should contact their state board or accrediting organization to determine if passing an exam after reading a book/magazine meets their CPE requirements. State CPE Sponsor #:_______________.How the New Leases Standard May Impact Business ValuationsBy Judith H. O’Dell, CPA, CVA1. The new leases standard will be effective for private companies in:a.Fiscal years beginning after December 15, 2018b. It is in effect nowc.Fiscal years beginning after December 15, 2019d. December 15, 20192. A lease is classified as a finance lease if:a.It transfers ownership of the underlying asset to the lessee by the end of the lease termb. The lease term is for the major part of the remaining economic life of the underlying assetc.The underlying asset is of such a specialized nature that it is expected to have no alternative use to the lessor at the end of the lease termd. All of the above3. After the effective date of the standard, the initial accounting by a lessee for a new lease is:a.Recognition of a lease liability at the present value of the lease payments discounted using the LIBOR rate and a right of use asset equal to lease liabilityb. Recognition of the right of use asset as the total cost of the lease and a lease liability in the same amount.c.Recognition of a lease liability at the present value of the lease payments discounted using the discount rate for the lease and a right of use asset equal to the lease liabilityd. Recognition of an asset equal to the value of item leased and a like liabilityVisit and log in to access an exam. Online exams are available for The Value Examiner issues from 2014 to current. You will be able to purchase, complete, and earn five hours of NACVA CPE* for each exam. You will instantly receive a certificate of completion for each exam you pass.Earn CPE Online by Reading The Value Examiner®!* This exam does not qualify for NASBA QAS CPE credit. Individuals should contact their state board or accrediting organization to determine requirements for acceptance of CPE credit.To learn more, please visit The Value Examiner CPE exam can now be taken online! CPEexamVEad.indd 110/11/16 1:01 PMA PROFESSIONAL DEVELOPMENT JOURNAL for the CONSULTING DISCIPLINES12 JANUARY | FEBRUARY 2019 t h e v a l u e e x a m i n e rSecurity Will be the Main Source of InnovationBlockchain allows secure transactions and prevents frauds, which is why blockchain trade finance has gained huge adoption, and the number of blockchain-based applications is increasing considerably. Given that security is the primary concern in the crypto economy, it will be the prime source of innovation in the crypto industry.Companies leverage decentralized biometric blockchain authentication as well as other blockchain-based apps to reduce data breaches and to provide more security. More and more businesses are expected to embrace the same solutions to deliver more secure and better services to their customers.CONCLUSIONWhere the cryptocurrency/blockchain/ICO markets take us from here may end up being more about discovering and measuring true economic value than the irrational exuberance that seemed to drive the Bitcoin price spike in late 2017. For valuation professionals, thinking about the fundamental value of Bitcoin by understanding the cost of the mining inputs as compared to the expected demand for the coins that are generated, discounted at a rate determined by a level of risk may look like a standard discounted cash flow model after all.Neil Beaton, CPA, ABV, CFA, ASA, is a Managing Director with Alvarez & Marsal Valuation Services in Seattle. He specializes in the valuation of public and privately held businesses and intangible assets for purposes of litigation support, mergers and acquisitions, incentive stock options, and estate planning and taxation.Mr. Beaton has written extensively on, and specializes in, the valuation of early-stage and venture-backed companies. His book, The Valuation of Early Stage and Venture-Backed Companies (J. Wiley & Sons, NY, NY) was published in 2010. E-mail: nbeaton@alverezmarsal.comVEMINING FOR CRYPTOCURRENCY1In the early days of cryptocurrency, anyone with a computer and a particular IT mindset could “mine” Bitcoin and other currencies. Today, to be profitable, most mining operations use specialized hardware called ASICs, which cost approximately $3,000 per rig for the current generation of hardware. In addition to the physical hardware, miners should understand blockchain and the cost of the inputs, specifically electricity. Cryptocurrency was designed to be decentralized, secure, and unalterable. Therefore, every transaction is encrypted. Once that encrypted transaction happens, it is added to a digital ledger using a “block” until a fixed number of transactions has been recorded. That block then gets attached to a chain—the blockchain—which is publicly available. Because privacy is one of the pillars of cryptocurrency, these transactions leave no trace of who is behind them. The location of the transactions is not centralized, either, so that it cannot be manipulated or controlled by one person or entity. Miners also verify the transactions, ensure they are not false, and keep the infrastructure running, all of which is built-in to the Bitcoin protocol.In simplest terms, once the software is loaded, a miner needs to mine a specific coin and edit an executable text file with details like the mining pool’s URL to connect to, the miner’s wallet address, and the name of the miner’s “worker” or PC. More advanced options allow miners to adjust how hard their GPU or CPU works. The vast majority of this software works across Windows and Linux, although it is more difficult to configure on non-Windows systems. What makes it more challenging is that these variables are formatted differently depending on the pools and the software.Once the blocks are solved, and coins are generated, the pool automatically pays the miners in the form of the underlying cryptocurrency directly to their wallet, or to an online cryptocurrency exchange that holds many different types of coins.For a deeper discussion and tutorial sessions, go to: mining-101- wha t -e xac t ly -i s -c r y pt o c ur r enc y -mining/#da840ada83a1 A PROFESSIONAL DEVELOPMENT JOURNAL for the CONSULTING DISCIPLINESVALUATION///////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////t h e v a l u e e x a m i n e r JANUARY | FEBRUARY 2019 13Editor’s Note: The study was completed in the U.K. market using prior research from the U.S. market, i.e., Blume and Vasicek. Since the U.K market has existed since 1801, it is safe to conclude that it is not only viable but advanced. In our global world, it is evident that finance theory is also global. This would mean that the result of this study can be applied to good effect in other advanced markets including in the U.S. The U.K. has about 2600 companies listed across a number of “markets.” The companies used in this study are those which are fully listed, and which are constituents of the FTSE All-Share Index—excluding financial sectors—hence 344. While the U.S. market is much larger, and this study was focused on the U.K., the results obtained are still statistically significant.A fundamental driver of the cost of capital is the systematic risk measured by the Beta coefficient. To achieve an accurate value of the cost of capital, the Beta estimates are required to exhibit a set of desired characteristics. Forecasted Betas should be unbiased, stable, and, most importantly, good predictors for future returns. The standard Beta estimation technique is the Ordinary Least Squares Regression (OLS) of security returns against market returns over four or five years of monthly data. However, there is strong empirical evidence of the limited ability of OLS Beta to meet the desired characteristics. Previous research reveals considerable debate about the appropriate alternative techniques. This current analysis examines six Beta estimation techniques in relation to three essential criteria: bias, stability, and predictive ability. BETA IN PRACTICEDespite the vociferous and sometimes fractious scholarly controversy surrounding its use and effectiveness, the CAPM Beta is still widely used as a measure of systematic risk in business valuation, financial performance evaluation, and the fair pricing of securities. Considered by many to be the most relevant measure of risk, CAPM Beta has proved to be extremely useful in business appraisal practice in the process of modelling an appropriate cost of capital. Within this, an assessment of risk impacting upon the business being valued is necessary. This can be modelled using a public company market proxy and the CAPM theory. A key conceptual constant here is the idea of using a stock’s historic CAPM Beta as a potential measure of future systematic or systemic risk. Beta, in this case, is a measure of a stock’s volatility against the overall market’s volatility excluding the unsystematic or specific risk. Systematic risk can be defined as the uncertainty of future returns due to uncontrollable movements in the market. This type of risk generally arises from external, macroeconomic factors that affect, to a greater or lesser degree, all economic assets within the overall economy. Whereas, diversifiable or unsystematic risk is based on firm specific factors. The values of historical Beta are used as estimates of the true future Beta for a stock and, as with all predictions, they are subject to errors and limitations. Therefore, the estimate of Beta may not be equal to the subsequently realized Beta in the future analyzed period. The process is further complicated as Beta is not perfectly stationary over time—as the fundamental characteristics of a firm change, Beta changes. The practical By Diana Raicov, MSc and Richard Trafford, MSc, FCT, CVA, CFE, MAFFVasicek and Blume Betas: Back to the Future (Part I of II)///////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////“We’ve already agreed that having information about the future can be extremely dangerous.” Dr. Emmett Brown (Back to the Future, 1985)A PROFESSIONAL DEVELOPMENT JOURNAL for the CONSULTING DISCIPLINES14 JANUARY | FEBRUARY 2019 t h e v a l u e e x a m i n e rviability and importance of Beta as a risk parameter is directly related to its stationarity, stability, explanatory power, and increasing reliability. Early examination of the stationarity of Beta coefficients carried out by Blume1 and Levy2 suggested that individual stock Betas estimated by the OLS method were not stable across the analyzed periods. However, the OLS method assumes that the Beta coefficient is stationary, and for this reason it might be considered a restrictive and inappropriate method. The stationarity problem was further approached in a number of studies which investigated the instability of Beta coefficients. Levy3 provides evidence of the stationarity of Beta in large portfolios. However, he admits that for smaller portfolios, Beta is unstable and even unpredictable in the case of individual securities. In contrast, Fabozzi and Francis4 in a study on the NYSE, conclude that Beta is a random coefficient. Their findings imply that less than a half of the total risk on NYSE stocks is explained by the market forces. The main reason for this is that the true Beta is moving randomly while the OLS Beta is a point estimate which is stable over time. In a previous test, Klemkosky and Martin5 also provided evidence of randomness in their study of Beta stability. Therefore, the use of the OLS method to estimate systematic risk is proven to lack efficiency due to the variability of the Beta coefficient. According to Chen,6 the OLS method will overestimate the portfolio residual risk. To mitigate the effect of Beta variability, various estimation adjustment techniques, such as the Blume and Vasicek adjustments, have been proposed.THE BLUME ADJUSTMENTBlume was the first to propose a technique to adjust Betas for their non-stationarity. He calculated Betas for 415 different companies and then formed portfolios of securities 1 Blume, M. E. (1971). On the Assessment of Risk. The Journal of Finance, 26 (1), pp. 1–10.2 Levy, R. A. (1971). On the Short-Term Stationarity of Beta Coefficients. Financial Analysts Journal, 27 (6), pp. 55–62. 3 Ibid.4 Fabozzi, F. J. and Francis, J. C. (1978). Beta as a Random Coefficient. Journal of Financial and Quantitative Analysis, 13 (1), pp. 101–116.5 Klemkosky, R. C. and Martin, J. D. (1975). The Adjustment of Beta Forecasts. The Journal of Finance, 30 (4), pp. 1123–1128.6 Chen, S. N. (1981). Beta Non-stationarity, Portfolio Residual Risk and Diversification. Journal of Financial and Quantitative Analysis, 16 (1), pp. 95–111.ranked by Beta. In a direct approach of the non-stationarity phenomenon, he observed that in large portfolios, formed by ranking individual Betas from lowest to highest, there is a tendency for the portfolio Betas to regress toward the overall mean of unity. In other words, Blume noticed that over time, the high Betas decreased and the low Betas increased. Therefore, he concluded that adjusting Betas toward unity improved their forecasting ability.Several reasons that explain this phenomenon have been proposed. Firstly, the tendency of regression to the mean might be caused by the real non-stationarities of Betas of individual securities. A frequently given intuitive explanation states that Beta’s instability might be caused by unstated economic or behavioral factors. For instance, in the case of high-risk firms, the risk of existing projects might become less extreme over time. Moreover, the new projects undertaken by the firm may imply lower risk than the existing projects. Therefore, if managers consciously undertake new projects to diversify, firm’s Beta will move toward the grand mean. Furthermore, firms with a low Beta exhibiting low systematic risk might be able to support more debt financing. The higher gearing will determine an increase in Beta toward one.A second reason identified by Blume relates to a statistical phenomenon called selection bias. This implies that some of the low or high ranked Betas obtained their rankings due to simple statistical under- or over-estimation in the ranking period. Therefore, Beta estimates are potentially contaminated by significant measurement error. However, in his paper, Blume proves that the selection bias is insignificant. Thus, Beta’s regression to the mean over time must be due to real non-stationarities. Consequently, Blume suggests that the mean-reversion tendency is reasoned by firms’ actual project selection.After identifying the tendency of Betas to move around the mean, Blume proposes a scheme of modifying Betas to capture this trend. To assess this tendency, the estimated values of Betas in one period have been regressed against the estimated values of Betas in the previous period. Therefore, the output was a line that measures the tendency of the forecasted Betas to be closer to one, which is the grand mean of all Betas over time, rather than the estimates from historical data. For the seven periods of seven years each analyzed starting from 1926 and ending in 1968, Blume obtained the following relationship between the two Beta estimates:A PROFESSIONAL DEVELOPMENT JOURNAL for the CONSULTING DISCIPLINESt h e v a l u e e x a m i n e r JANUARY | FEBRUARY 2019 15FIGURE 1: THE BLUME ADJUSTMENT FOR BETAThe equation in Figure 1 lowers the high values of Betas and raises the low values, bringing them all closer to one. Overall, the Blume adjustment is widely accepted as a way to reduce measurement errors in Beta estimates and is used by a number of leading data service providers including Bloomberg, Value Line, and Merrill Lynch. They weight the raw stock market Beta by a factor of 0.66 and the market portfolio value, which is one, by a factor of 0.33. However, these providers use only one of the many regression results obtained in Blume’s paper. Moreover, the Blume analysis may be considered outdated.7 Hence, other Beta adjustment techniques which have been developed later have been shown to outperform the Blume adjustment in terms of efficiency and accuracy.THE VASICEK ADJUSTMENTAn alternative moderation technique suggested by Vasicek,8 demonstrated in figure 2, is based on the property that in the forecast period, Beta tends to be closer to the average Beta than is the estimate provided by the historical data. Therefore, Vasicek implemented a straightforward technique to correct for this tendency by adjusting each Beta toward the average Beta. For instance, a simple way of achieving this would be to assign fifty percent weight to both historical and average Beta and add them. The result would be an adjusted Beta toward the mean. However, such a method should not adjust all the stocks in the same way but should consider the level of uncertainty regarding 7 Ogier, T., Rugman, J., and Spicer, L. (2004). The Real Cost of Capital: A Business Field Guide to Better Financial Decisions. Glasgow: Prentice Hall.8 Vasicek, O. (1973). A Note on Using Cross-sectional Information in Bayesian Estimation of Security Betas. The Journal of Finance, 28 (5), pp. 1233–1239.βί2 = 0.343 + 0.667 βί1Where:βί2 – Beta in the second (next) periodβί1 – Beta in the first periodthe individual Beta. The following scheme proposed by Vasicek moves each historical Beta toward the average and also considers the size of Beta sampling error.9FIGURE 2: THE VASICEK ADJUSTMENT FOR BETAThus, this weighting procedure adjusts observations with large standard errors further towards the mean than it adjusts observations with small standard errors.As Vasicek asserts in his paper, this is a Bayesian estimation technique which utilizes information prior to sampling as well as sample information to reach an optimal prediction. Therefore, the method weights knowledge which is available in addition to the sample information. This is essential when estimating Betas of stocks, as the prior information is usually substantial. In other words, the Vasicek adjustment produces an unbiased Beta estimate if a single OLS Beta estimate is observed and there is information regarding a prior distribution. Consequently, it is more appealing than the Blume adjustment due to the weights placed on OLS Beta. They are practically the result of a reasonable economic rationale which asserts that the estimation error leads to biased OLS estimates of true systematic risk. Thus, the correction proposed by Vasicek is based on the precision of the sample estimate.In addition, the Bayesian technique, as proposed by Vasicek, does not imply a trend in Betas which the Blume technique does. However, another source of bias which might affect it is related to high-Beta 9 Elton, J. E., Gruber, M. J., Brown, S. J., and Goetzmann, W. N. (2011). Modern Portfolio Theory and Investment Analysis (8th Ed.). Chichester: John Wiley & Sons Inc.1-xo βί1 2o β–1 2 + o βί1 2o β–1 2o β–1 2 + o βί1β–1 +βί1 βVasicek = Where:βVasicek – Beta under Vasicek Adjustmentβί1 – Beta in the previous periodβ–1– Average Beta across the sample of stocks in the historical periodo βί1 2 – The squared standard error of the estimate of Beta for security i in the previous time periodo β–1 2 – The variance of the distribution of the historical estimates of Beta over the sample of stocksxA PROFESSIONAL DEVELOPMENT JOURNAL for the CONSULTING DISCIPLINES16 JANUARY | FEBRUARY 2019 t h e v a l u e e x a m i n e rstocks having larger standard error than low-Beta stocks.10 As the shift in the Vasicek technique is dependent upon the standard error of the sample Beta estimate, high-Beta stocks will have their Betas lowered by a greater percentage than the percentage increase for low-Beta stocks. Thus, the average of the estimate Betas will actually be lower than the average historical Beta and the adjustment brings a decreasing trend in Betas. This property brings bias in the estimation if Betas will not continually decrease. Therefore, further improvement can be made by adjusting all Betas marginally upward, so they have, overall, the same average as the ones in the historical period.INTRODUCTION TO THE CURRENT STUDYIn terms of ability to forecast, both Blume and Vasicek techniques have been extensively tested in the literature. As expected, both techniques achieve more accurate forecasts of future Betas than the unadjusted Beta. However, when compared, the Vasicek technique proved to slightly outperform the Blume technique. Klemkosky and Martin11 evaluated these two methods of adjusting Beta by their ability to better forecast Beta. Their conclusion downplays the Blume technique in favor of Vasicek’s. The primary focus of this current study is on the comparative performance of the Vasicek Beta estimates in the U.K. market. Three criteria will be considered in order to reach a conclusion (see Figure 3) regarding its ability to provide an accurate forecast for Betas: bias, stability, and predictive ability. This approach has been previously employed by Gray, Hall, Klease, and McCrystal12 in a study on the measurement of systematic risk in the Australian market.FIGURE 3: BETA ESTIMATES—CRITERIA OF APPRAISALUnbiasedness is a desired characteristic of Beta and implies that the expected return provided by the CAPM has an equal chance of over- or under-estimating realized returns. In other words, the OLS Betas are generally considered unbiased if the expected value of the sample equals the population parameter: E (β) = β. 10 Ibid.11 Klemkosky, R. C. and Martin, J. D. (1975). The Adjustment of Beta Forecasts. The Journal of Finance,I (4), pp. 1123–1128.12 Gray, S., Hall, J., Klease, D., and McCrystal, A. (2009). Bias, stability, and predictive ability in the measurement of systematic risk. Accounting Research Journal, 22 (3), pp. 220–236.Consequently, Beta estimates exhibit bias which, in the statistical sense, means that if we drew repeated samples of stock, there is an uneven chance that an additional draw would generate a Beta estimate above or below the OLS estimate. However, this statistical definition is useless when the true population Beta is unknown. The reason is given by the fact that the original sample might not be representative of the true Beta. Thus, practitioners require an unbiased Beta estimate that ensures the true population Beta has an equal probability of lying above or below it. Hence, the Beta adjustment techniques need to be assessed on their ability to mitigate against bias. Consequently, in this research, the mitigation of bias will be one of the three criteria for evaluating the Beta adjustment techniques. The second criterion employed is the stationarity of Beta estimates. There is analytical and empirical evidence that OLS Beta estimates are volatile, which has been shown to result from the fluctuations of firms’ systematic risk. Acquisitions, investments, and leverage changes represent only some of the key drivers of systematic risk variability. However, the volatility of OLS Beta estimates is thought to exceed what can reasonably be explained by these changes.13 Apparently, the length of the estimation period has an important impact on the stability of estimated Betas. Baesel14 concluded that the stability of Beta estimates improves as the length of the estimation period increases. When predicting rates of returns, there is a necessity for Beta estimates to be stable. Unstable Betas would contribute to biased perceptions regarding risk and returns. The third essential characteristic of Beta estimation is its ability to predict the returns. Therefore, the stronger the association between the adjusted estimates for Beta and the future returns, the more accurate and realistic are the estimates. This argument is consistent with Rosenberg and Guy’s paper15 which attests that Betas should be assessed 13 Ibid.14 Baesel, J. B. (1974). On the Assessment of Risk: Further Considerations. The Journal of Finance, 29 (5), pp. 1491–1494.15 Rosenberg, B. and Guy, J. (1995). Prediction of Beta from Investment Fundamentals. Financial Analysts Journal, 51 (1), pp. 1975–1984.Criteria for evaluating the performance of Vasicek beta adjustmentsUnbiasednessStabilityPredictiveAbilityA PROFESSIONAL DEVELOPMENT JOURNAL for the CONSULTING DISCIPLINESt h e v a l u e e x a m i n e r JANUARY | FEBRUARY 2019 17on their ability to predict future stock returns. Consequently, returns predictability is considered a fundamental criterion in the assessment of Beta and the adjustment techniques. FILTERING MECHANISMSApart from the mean reversion and selection bias adjustment, further filtering or winsorizing methods will be examined. These techniques have been extensively used in practice with the aim to exclude less precise estimates from the samples. Therefore, their use might increase stability and returns predictability. However, the filtering methods might also introduce bias in calculations, so their use involves a trade-off between increasing precision and accepting a side-effect. FIGURE 4: FILTERING ADJUSTMENTSAs shown in Figure 4, the adjustments are to exclude: •Firms with a market capitalization of less than 100 million British pounds (GBP)Alternatively, to exclude firms where the regression artefacts show low explanatory power (i.e., an R2 less than ten percent) •Firms with a low result validity (i.e., t-stat less than two)METHODOLOGY AND DATAThe performance of alternative estimation techniques has been scrutinized in a number of studies. A common approach of examining Beta estimates is to evaluate their ability to forecast subsequent period Betas. This paper considers the performance of Beta estimation techniques (see Figure 5) in the U.K. market, on an industry basis. Three criteria have been employed in assessing the estimation techniques: level of bias, stability, and return predictability. The main objective is to find the best method for estimating Beta in the context of the three chosen criteria. With this purpose, the study seeks to compare the six techniques involved in Beta estimation against these criteria to reach a ranked conclusion regarding the best method.CAPM theory tells us that the representation of the overall market should include all asset classes, but the calculation is generally made tractable by using a market index to represent the whole market. The index which the individual stock is then regressed against will be dependent upon where a business conducts the major part of its economic activity. In the U.S., this may be the S&P 500 or the NYSE index; in the U.K., the FTSE-All-Share; and in Germany, the CDAX; and so on. In addition, Beta is measured using a periodic timescale of sufficient length to incorporate a broad cycle of economic events. The periodic timescale varies from user to user: for example, Value Line uses weekly stock returns over a five-year period regressed against the NYSE index movement; Compustat uses monthly stock returns again over a five-year period regressed against the S&P 500. Bloomberg provide a few Beta calculations using daily, weekly, or monthly stock return data. The analysis in this work has been carried out by using a sample which comprises of twenty years, 1991 to 2011, of monthly returns on the U.K. FTSE All Share Index and all the companies comprising it. The monthly returns data for the whole sample period have been sourced from Datastream (Thomson Reuters database). The assessment is conducted on an industry basis and the companies have been sorted by industries according to the four-digit Global Industry Classification Standards (GICS) Code, which, in recent research, has been shown to be the most effective industry classification system.16 The resulting sample includes twenty industries, excluding, as is standard practice, the financial sectors. Six variant industry Betas have been forecasted by applying the following estimation techniques. These techniques produce six different estimates of Beta for each company on a monthly basis in the analyzed period. 16 Casey, R. and Schaberl, P. (2017). A Recent Comparison of Industry Classification Schemes Using Publicly Traded Firms. The Value Examiner, January/February, pp. 6–13.Small size firmsMarket capitalisation under ₤ 100 millionR2 <10%Firms where the returns' regression exhibitst-stat<2ExcludeA PROFESSIONAL DEVELOPMENT JOURNAL for the CONSULTING DISCIPLINES18 JANUARY | FEBRUARY 2019 t h e v a l u e e x a m i n e rFIGURE 5: BETA ESTIMATION TECHNIQUESThe OLS, Blume, and Vasicek Beta estimates have been computed, replicating the study by Gray et al.,17 by using the monthly returns from the prior four years (forty-eight months). Therefore, the twenty years period is split into five sub periods of four years (forty-eight months) ending: 1995, 1999, 2003, 2007, and 2011. Each Beta estimate is calculated for each of the five periods using the previous forty-eight months of historical stock returns. The results are used to analyze the performance of the alternative techniques in different time frames. Firstly, the OLS Beta representing the coefficient of regression of security returns against market returns is calculated. Next, a Blume adjustment has been attained as it accounts for the tendency of Betas to regress towards the absolute average of the market. The computation of the Blume adjustment employs the formula used by global data service providers such as Bloomberg or Value Line. Therefore, the coefficients placed on the OLS estimates (0.33 and 0.67) are held constant. The second adjustment applied to the OLS figures for Beta estimates is the Vasicek technique. The model employed generates a Beta estimate which is a weighted average of the sample and a prior prediction of Beta drawn from the population. Therefore, the method shifts Beta towards the mean while considering its standard error in the previous period. The size of the adjustment is proportional to the uncertainty regarding the prior estimations. Thus, the final result depends on the 17 Gray, S., Hall, J., Klease, D., and McCrystal, A. (2009). Bias, stability, and predictive ability in the measurement of systematic risk. Accounting Research Journal, 22 (3), pp. 220–236.OLSR2>0.1t-stat>2R2>0.1& Mkt. Cap > ₤ 100 mBlumeVasicek • Result from the regression of stock returns against market using 20-28 months of returns •β= • Mean OLS Beta that only includes firms where the regression has a R squared above 10% • Mean OLS Beta that only includes firms whose t-statistics is greater than 2 • Mean OLS Beta that only includes firms with an R squared above 10% and a market capitalization which exceeds ₤ 100 m •β Blume= 0.33 +0.67* β OLSo imo m2o βί1 2o β–1 2 + o βί1 2o β–1 2o β–1 2 + o βί1β–1 +βί1 βVasicek = A PROFESSIONAL DEVELOPMENT JOURNAL for the CONSULTING DISCIPLINESt h e v a l u e e x a m i n e r JANUARY | FEBRUARY 2019 19population mean and standard deviation. However, these two parameters are not observable, but they are estimated from the distribution of OLS Betas on the U.K. Stock Market. The calculation of Vasicek Betas is described in Figure 6.FIGURE 6: THE CALCULATION OF VASICEK BETASThus, the Vasicek adjustment has been applied to the OLS Betas reported at the end of each estimation period. The second parameter in the formula is the mean OLS Beta which represents the four-year (forty-eight months) average Beta for the whole sample. Next, for the calculation of squared standard error, the typical statistic formula has been used. After obtaining the three different forms of Beta estimate, raw OLS, and OLS Beta adjusted using the Blume and Vasicek models, descriptive statistics for Beta estimates have been reported at 31 December 2011. This highlights the level of risk per industries and provides an overview of the U.K. market context. The results display an indication of the variation of Beta estimates at a static point in time. Hence, the three primary estimation techniques considered (raw OLS, Blume, and Vasicek) deliver a certain mean Beta and its standard deviation for each industry. The descriptive statistics highlight both equally-weighted and market capitalization-weighted mean Betas for the three estimation techniques. In addition, the industries are divided into high, average, and low risk industries, and the descriptive statistics are provided for each of them.o βί 2o β–1 2 + o βί 2o β–1 2o β–1 2 + o βί 2β–1 +βί βVasicek = βί – OSL Beta at the end of the 5 estimation periods - i =December 1995, 1999, 2003, 2007, 2011βί – the mean OLS Beta for each periodo β–i 2 – squared standard error calculated by using the formula:o βί 2 = a = Ri – βί *Rm,where Ri is the average security return for every 48 months and Rm is the average market return for every 48 monthso β–1 2 – the variance of OLS Betas in the prior period1-xxΣ(Ri – a – βί*)2/(48-2)48i = 1Market variance * (48–1)Next >