Comparative_valuation_methods_fact_and_f.docx

Comparative Valuation Methods: Fact and Fallacy of Dhaka Stock Exchange Mohan L Roy Department of Business Administration, IBAIS University, Bangladesh *Corresponding Author: [email protected], [email protected]

Copyright©2016 by author, all rights reserved. Author agrees that this article remains permanently open access under the terms of the Creative Commons Attribution License 4.0 International License

Abstract: This survey investigates how far popular theories and models often used in pricing common stocks are pragmatic in context emerging Bangladesh economy where public perceptions are inarticulate. The study presents an approach selecting analogous companies in market-based research and equity share valuation. Steered by valuation theory, this study initiated a ‘warranted multiple’ for each stock, and categorize peer group having the closest nature of business, capitalization, earnings and growth. The adopted tests cover the broad-spectrum of stocks data sampled from financial service companies: 2007-2015. The approach tested the efficacy of relative models in foretelling future P/E and P/NAV, and concludes that all predictions are significantly diverse from the actual prices and proves ‘warranted models’ are ineffective.

Keywords: Relative Price, Intrinsic Value, Emerging Economy, Price-Earnings multiples, P/NAV

1. Introduction Illiquidity, ineptitude and irregularity of Bangladesh stock markets (BSX) is wiping out gradually. This trend is synchronized most possibly because of growing add up to stock exchanges lists and size, and government market oriented initiatives. In global economies, the typical markets are called ‘emerging’. Albeit each of these markets has self-style feature, studies [1] categorize them into three familiar phases. First, the typical stock markets incline to building up by progressive macro policies which eventually possible if the country would accomplish certain degree of socio-economic steadiness by boosting up investors’ confidence for market participation. In second phase, degree of credibility rises as market liquidity grows and risk-return trade off close up, and cross border funds mobilize more towards other global markets owing to higher potential gains. The third phase entices extensive development and takes away vulnerable returns that consequently cheer investors for more IPOs. In the end the market risk premium closes to

risk free rate that blows up to unwavering maturity phase, considered ultimate stage. Because of this market (BSX) self-style nature, it is hard to discern the exact phase where the market lies. The panorama of this market is unmatched not only in typical Asian markets, also in global emerging economies. Profitability, growth, and its’ great set back experiences (market crash 1996, 2011) are different from others. However, buoyancy about stock market has been changing positively that drive mounting size of participants, turnover, number of listed companies and market capitalization. Studies discovered [2] responsible factors behind such steady progress of the BSX are primarily GOB initiated goal oriented policies and programs. In view of such assumption, the BSX may be aligned in the second phase.

2. Review of Literatures Today one of the fundamental issues is how to value common stock? Which method has in times gone by performed the best? The investors have many alternative measures for valuation but mostly rely on price-to-earnings ratio (P/E), price to book (P/NAV), the enterprise value to earnings before interest, taxes, depreciation, and amortization (EV/EBITDA) ratio and some other basic facts like dividend declaration, right issues, trade volume etc. For the time being, conventional studies [3] emphasized on the price to book ratio (P/NAV), and recently more researches [4] brought in gross-profits measure (GP).Valuation measures which incorporate last year's earnings or forward earnings are interesting, but what about long-term valuation measures? As early as 1930s, the market analysts and investors uphold average earnings, not less than at least five years, instead of merely twelve months approximation to envisage stock price. Prior study [5] suggests using current earnings in the context of valuation metrics, and earnings in P/E, "should cover a period of not less than five years, and preferably seven to ten years.” Later in 1998, more academics [6] recommend that annual earnings are strident as a measure of basic value. In recent times, some others [7] conducted a full-bodied study of long-term P/E ratios (U.K.

stock market data: 1975-2003) and found support that using a long-term earnings average (eight years) instead a single year boost up the spread in returns between value and growth stocks by more than five percents. Also another prominent research found the similar result and supports the prior conclusion [8]. In contrast, some researches evidenced that longer-term (i.e., less than 8 years) metrics are not reliably better at predicting returns than one year metrics [9]. Nevertheless, this study is incapable to imitate such results in Bangladesh stock markets and finds negative results with long-term valuation measures. In 1961 the seminal work [10] unanimously known M&M view, raised a basic question: “What does the market ‘really’ capitalize?” The view did not crown a winner among approaches that rely on earnings, dividends, or cash flows. Rather, it presented that all these methods generate almost indifferent inference if the solver can set problem properly. Nevertheless, numerous approaches have been implied over the decades but none accomplished perfection 100 percent on indifferent result. Despite so, among the approaches the earnings multiple enveloped recognition as easiest and primary method which analysts, investors often use to value common stocks. Yet, most users do not have comprehensible idea of what a particular multiple implies about a company’s future financial performance and do not understand how multiples change over time [11]. Determining stock value, multifaceted and animated processes, requires inputs of numerous estimations perceivably less reliable in emerging markets in light of market efficacy [12]. Using such oversights in pricing of stocks thus often generate return unusual. The investors lose or gain frequently due to dim forecasting matrices of target stock prices which eventually meet mispricing. Translating forecasting matrices into valuations using different ‘absolute and relative’ valuation methods is convention globally even though they differ frequently [13]. Absolute valuation models, such as discounted dividend model (DDM) and DCF models rely modestly on data on the company being analyzed. The value of company asset is the present value of future cash flows. A good DCF model avoids accounting vagaries, whereas managements can manage or manipulate earnings [14]. The reporting of earnings and financial position of a company involves both significant independence in selecting from the sphere of GAAP and inevitable need for management to make numerous valuation estimates and judgments [15]. Such flexibility in presenting outcomes help the company management portraying financial affairs in the best impressive possible is a common phenomenon. Relative valuation models, such as P/E, price-to-book ratio, and price-to-sales ratio, seek to determine the true value of a stock by comparing the multiple of the company being analyzed with those of similar companies. In the past, analysts opt to yardstick the valuation of the studied stocks by selecting companies of like size and identical industry. Being the selection of these comparable companies has often been subjective, and the inclusions of improper companies led to big valuation errors [16]. The P/E multiple remains the

primary method analysts use to value stocks [17] and equity valuation researches being surveyed found that more than ninety-nine percent of the analysts used some sort of multiple and less than thirteen percent used any variation of a DCF model [18]. P/E multiples may be a common way to assess the attractiveness of a stock, but most investors fail to have a clear sense of what a particular multiple implies about a company’s future financial performance and do not comprehend how multiples change over time[19]. More researches argued, as value determiner the multiplier models may lead to misleading notion. Some of them critically acclaim that: ‘valuation is a fallible subject’ and any method will result in a misleading prediction at sometime [20]. An early 20th century economist [21], (Milton Friedman) expressed: “price is what you pay, value is what you get” and also pointed out that “investors need a valuation model to understand value and for a valuation model it is a way of understanding a business and how it generates values”. This pioneer economist probably meant typical relative valuation methods that synchronizes ratio for comparing the sample companies’ prices with some sort of monetary value. Later advancement introduced numerous factors like EPS, net book value per share (NAV), earning before depreciation and taxes and dividend per share (DPS) or cash flow often screening price of companies’ stock [22-27]. These fundamental factors meticulously are expected to reflect market price if they are adjusted authentically. Because of their comparative simplicity for adjustment, these ratios are most widely used in emerging markets like Bangladesh [28]. Survey found, many analysts use these comparatives frequently for examining peer group stocks based on relatively low ratio believably that companies carry ‘sound business health’, and less volatile relative to companies with higher ratio [29-31]. Considerable evidence suggests that such low multiples are inversely related to higher constant future return [32]. Supporting the recommendations further observation evidenced [33] that the lower is the multiple the higher is the valuable for predicating future returns. Several studies beginning with Basu (1977) [34] have examined the relationship between historical P/E ratio for stocks and return of the stocks. Some have suggested low P/E ratio stocks will outperform high P/E ratio stocks because growth companies enjoy high P/E ratios, but market tends to overestimate growth potential thus overvalues these growth companies while undervaluing low-growth companies with low P/E ratios [35]. Another contemporary research [36] examined adjusted low P/E ratio for firm size, industry effect and infrequent trading and likewise found that the risk-adjusted returns for stocks in the lowest P/E ratio quintile were superior to those in the highest P/E ratio quintile. However a widespread disapproval of all these multiples is that they do not consider the future, and are often computed from trailing history for the current values of the divisor [29].The economic irrational undermines the method of comparables is the law of one price [29] used as denominator of the multiples for future prediction. But the financial market trend repeats history is myths that barely happen in economies which emerge for development.

3. Research Questions A common view is that whatever market type (volume, volatility, reliability of data, industry concentration etc.,) the relative methods perceive the correct price of common stock. The techniques set logical link between earnings, prices and NAV for valuing equity stocks in portfolio. Now question: Is it fact or fallacy? This methodological proponent is really effective in emerging market like BSX?

4. Methodology 4.1. Sample Stocks, Data Variables and Parameters Using common valuation methods numerous initiatives [37-39] carried out studies in different economies and times, and found dissimilar results. We sampled 39 stocks of financial service industry from DSE websites to test the efficacy of the relative valuation methods in BSX. Variable data like EPS, NAV, dividend, and PE ratios have been taken covering period 2007-2015. Identifying peer group for ‘comparable’ was tedious and thus finding comparables, stocks have been sampled from the pool of similar line of business and industry having comparable capitalization, profitability and growth1. Data materiality was checked by simultaneous examination of respective CARs. The stocks are sampled and tabulated separately showing 14 banks of 30, 12 NBFI of 23 and 13 Insurance companies of 46 (population)2 respectively. The warranted models (M1) and (M2) forecast prices P1 (CBV-company PE based value) and P2 (GBV-Group PE based value) for each stock under three sub group respectively. 4.2. Synthesized Models–the Value Determiner Warranted multiple models have been formed to predict prices comparable to prices in market during target period 3. Using valuation theory, the researcher adds objectivity to this process while improving the validity of the relative valuation models. This differs with the traditional use of relative valuation models on ground that criteria for the warranted multiple include the multiple of a closely matched individual stock, or the average the multiples for the peer group. This approach is based on the belief that companies are truly comparable if they have similar fundamentals4. In addition, although previous studies assume that market price is reflected correct price, the warranted multiple is allowed to deviate from the market multiple, and the efficient market hypothecation (EMH) is thus not necessary to matrix the models [16]. Weighing merits to this method the few studies

1 Growth (g) = (1/PE Ratio)*(1- D/P ratio); Brealey& Myers: Principles of Corporate Finance, 6/e p.315 2 46 Insurance companies are listed, but only 28 companies disclosed required data every year of the sample period. 3 Immediate last six months: August 2015- January 2016 4 Profitability, growth, P/NAV and PE ratio

[34, 40-43] favored its potential fitting with P/NAV (Price to book value) ratios to avoid imitation in EPS often conveyed by fuzzy companies. Comparative investment risk, past earnings and growth potential are likely factors often affect peoples’ mind are reflected in P/NAV ratio correctly but less likely in PE multiples [16]. The warranted models thus adjust risk-return and growth necessarily absorbing exposures to PE ratio. Model 1:8-year trailed (2007-14) company average P/E multiples * 1- year (2015-16) forecasted EPSt+1 CBVj.t+1 =P1= P/Ej.t * Rj,t+1……………(M1) Model 2:8-year trailed (2007-14) peer group average P/E multiples * 1- year (2015-16) forecasted EPSt+1. GBVj.t+1=P2= P/Ej. t≈i* (Rj,t+1)……………(M2) Where, CBVj.t+1refers predicted value of stock j at time t+1; P/Ej.t 8-year average trailed PE ratio of company j at time t; Rj.t+1stands1 year forecasted EPS of company j; subsequently model (M2) replaces company average PE by peer group average keeping all other parameters unchanged. 4.3. Formulation and Testing of Hypotheses Using valuation theory, this study adds objectivity to this course of action while improving the soundness of the comparative valuation methods. The argument lies behind that imitating profitability has least impact on equity value over the long term period. Temporary rise or fall in share prices might be occurred due to short term monetary reasons such as arbitraging, short selling etc. but not because of profitability. Second, the net asset value (NAV) is a residual figure after dividend and other appropriations, and the value is carried over till the end of next year. An unusual change in P/NAV ratio if observed during the period that believably would not sustain over the year. So short term periodic changes in price do not occur because of NAV but because of some invisible reasons. On the two propositions, we hypothesized that earning multiple and net asset value are most likely reflective on forthcoming share price. Accordingly this study sets two hypotheses: 1. Ho: Models estimated prices equal actual price (Po, 6-month average); i.e., (Po=P1=P2). H1: Significant divergence lies between the prices (Po≠ P1≠P2). 2. Ho: Estimated prices and NAV; i.e., (NAV=P1=P2). H1: Significant differences lie between them i.e., (NAV≠ P1≠P2). A non-parametric statistical analysis has been used to screen out number of stocks (frequencies in %) performed correctly (if value fall in 100±10% of actual price range), overvalued (if values go over 110%) and undervalued (if values fall below 90%). For inferential analysis, ANOVA: single factor (α=5%) has orchestrated the divergence between models determined values and market prices. The single factor experiment signifies differences in corresponding column values noted

by NAV (net asset value per share), AVP (average market price), CBV and GBV (forecasted by models M1&M2 respectively). All non-parametric and parametric analyses are shown in appendix tables 1.1& 1.2, and 2.1, 2.2, 2.3, 3.1, 3.2& 3.3 respectively. 4.4. Limitations and Caveats Only financial service companies were prime judgment for sample selection. Not all DSE listed companies have been considered. Authenticity of material public information is not unquestionable. Moreover DSE official website publication is not complete and time befitting, and thus up to date data (2015-16) has not been fed into equations. Computing growth rate, data published in the financial daily5 was taken into account which was not gone through cross examination due to unavailability of required CARs. Parametric test has been applied on ground that the profitability, NAV, PE ratio, P/NAV and DPS of the sampled companies are assumed approximately normal without having adequate evidences to hold up the argument.

5. Outcomes: Facts and Fallacies Fama and French hint that different price ratios are "pretty much as good as another for this job" of explaining returns. Though this suggestion came in different times (two decades before) and in different economic environment (not emerging economies), this review regretfully holds opposing views. A significant difference found lying in column values, and arguably the utterance came earlier is a myth, not a fact. Models’ effectiveness considerably depends on flows of material information in market which believably have an effect on share prices. Weighing critics to PE ratio trailed past economies, the underlying methods found fallacious. The rise and fall of price less likely response to such past economic environment even though market participants often rely on lagging indicator (PE ratio). Factually it can rarely meet people targets. Descriptive statistical analysis figured out frequencies in percents of stocks (table 1.1 &1.2) valued fairly by the models (M1&M2) separately for three sub groups. The findings suggest that ‘peer group’ is now a ‘buzz word’ but it does not bring sense in financial market of Bangladesh. It is merely an operative ‘trade word’ and has little impact on pragmatic practices. While ‘one peer group’ experiences down fall, the other non-comparable groups are often lining up in the same row but the opposite scenario is often observed in the homogeneous group. The reason behind such uncharacteristic trend is still unfolded. Catering companies stocks for valuation practically is a waste of time, even though this research initiative has taken samples on such ground. Using company based PE multiples, 14.29 (banking) and 7.14 percents found fairly priced and underpriced respectively, and the rest are gone too overvaluation mode.

5 The Financial Express, Stock/Corporate, 31 December 2015, p.10

We found only 7 percents stocks (bank) valued fairly with the group average PE multiple and the rest went to overvaluation ranging from ten to above two hundred percents. Other two groups hold stock fairly valued zero percent by this model (M2) (table-1.2). Further with the group PE multiple only 8.33 percents shows fairly valued which hailed from NBFI group but none (zero percent) from other two groups (table-1.2). Nevertheless, of the selected samples price estimation of stocks in banking group resulted better off to other groups. Apparently all three peer categories exposed almost similar blueprint. Inferential statistical outcomes carried out by parametric test, ANOVA (single-factor), rejected both null hypotheses except the case for NBFI with underlying assumption of AVP = CBV= GBV [table 2.2]. The resultant F statistic 6.06 >F Critical 3.26 and P=.01(Banking), 16.17≥ 3.28 and P=0.00(Insurance) be evidence for no relation exists between the column values i.e., AVP ≠CBV≠ GBV. The method prove different view for NBFI where resultant F statistic1.92
6. Reprise and Conclusions The study pictures out the fallacies and facts of BSX where PE and P/NAV ratios exist long standing for estimating stock prices. Witnessing vulnerable equity share market returns over the decades people lost confidence on methods of what so ever-conventional or non-conventional. Weak regulations, incessant buoyancy, fragile integration among regulators, analysts, academics and researchers persistently made public material information less reflective of prices in markets. Bizarrely theoretical researches or models found no substantial development to discern causes behind unpredictable movement of prices in so called emerging markets like BSX. Financial service companies disclosures over the period 2007 to 2015 have been surveyed for: first EPS and market price mirror image thereon; 2ndly, the transform in NAV over the period and price reflection. The test cases have been arrayed in stocks of three separate groups for predictions which all proved infeasible. The correct value matrix of equity shares relied on comparatives is a myth; game theory is a fact. Making investment decisions based on PE and P/NAV multiples is difficult. The warranted PE multiples of peer group as well as individual company produce an absurd percent of acceptable forecast for stock values. Atypical outcomes tendered other peer group. The price movement over the sample period was so volatile that the differences with determined values found incongruous, and any decision for equity deal would be purely on game theory, not on fundamentals as the study observed.

A common criticism of all typical relative multiples is that the models mull over past for future prediction. Emerging economies hardly ever repeat the history and pass through spectacular changes by engrossing supernormal growth. This study fits growth6adjustment for each warranted multiple to defy this criticism. The resultant descriptive statistics clearly shows that warranted models calculate values utterly over the disclosed NAV and prices in market. Even though portraying market efficiency was not leading motive, the study initiated to explore common factors cause changes in people’s perception while dealing common stocks in markets. Apparently the visible changes in market conduct came in owing to automation, changes in regulatory and institutional laws, scaling of market indices, online market observation and control over market sharing. In question of rationality, the market participants and the market are so volatile that no formula can track down perfectly the traits of the commons and the market as well; despite so people are always keen to hear from the legislator, regulators pertinent to market mechanism. Equity stock prices are most likely influenced due to influx of new material public information, and also equally likely are affected owing to key person of GOB while they unleash remarks on the share market issue. Empirical observation evidences that as many as made investment because of good fundamentals, far more let out investment funds due to other reasons of return. These market participants are deliberately either ignorant or least caring about fundamental issues. As a result ‘market efficiency’ is a myth, not a fact.

REFERENCES [1] M.G. Papaioannou and L.K.Duke, Internationalization of Emerging Equity Markets, Finance and Development, 1993 30(3) 36-39 [2] M. Farid Ahmed. A Conspicuous Growth in Stock Market, The Financial Express, (Daily news paper) Dhaka, 18 September (2014) [3] Fama, Eugene F., and Kenneth R. French, “The Cross-Section of Expected Stock Returns.” The Journal of Finance. 47 (1992), 427–465. [4] Novy-Marx, Robert, “The Other Side of Value: Good Growth and the Gross Profitability Premium,” Working Paper, University of Rochester, 2010. [5] Graham Benjamin, D. Dodd. Security Analysis, New York: McGraw-Hill, 1934. p. 452

Accounting. 37 (2006), 1063-1086. [8] Loughran, Tim, and J. Wellman, “New Evidence on the Relation Between the Enterprise Multiple and Average Stock Returns.” Journal of Financial and Quantitative Analysis 46 (2012), 1629-1650. [9] Gray, Wesley, and J. Vogel, “Analyzing Valuation Measures: A Performance Horse Race over the Past 40 Years.” Journal of Portfolio Management 39 (2012), 112-121. [10] Merton H. Miller and Franco Modigliani, "Dividend Policy, Growth, and the Valuation of Shares,” Journal of Business, Vol. 34, No. 4, October 1961, 411-433. [11] Michael J. Mauboussin, “M&M on Valuation,” Mauboussin on Strategy, January 14, 2005. [12] Stanley Block, “Methods of Valuation: Myths vs. Reality,” Journal of Investing, Vol. 19, No. 4, Winter 2010, 7-14. [13] Nikos Spiliopoulos. Valuation Model Choice and Target Price Accuracy of Equity Research Reports of Firms from the Utilities and Energy Sector, MSc Economics & Business Thesis, ERASMUS School of Economics, University of ROTTERDAM, (2011 p.5) [14] Paul Asquith, Michael B. Mikhail, and Andrea S. Au, “Information Content of Equity Analyst Reports,” Journal ofFinancial Economics, Vol. 75, No. 2, February 2005, 245-282. [15] A. Damodaran. Valuation Approaches and Metrics: A Survey of Theory and Evidences, Stern School of Business, November 2006 [16] Sanjeev Bhojraj and Charles M.C. Lee: Who Is My Peer? A Valuation-Based Approach to the Selection of Comparable Firms, Journal of Accounting Research vol. 40, no. 2 (May 2002):407–439 [17] Juliet Estridge and Barbara Lougee, “Measuring Free Cash Flows for Equity Valuation: Pitfalls and Possible Solutions,” Journal of Applied Finance, Vol. 19, No. 2, Spring 2007,6071. [18] Tim Koller, Marc Goedhart and David Wessels, “Valuation: Measuring and Managing the Value of Companies”, 5/e (Hoboken, NJ: John Wily and Sons (2010) 787-90. [19] Michael J. Mauboussin and Dan Callahan: What Does Price-Earnings Multiple Mean? An Analytical Bridge between P/Es and Solid Economics, Credit Suisse, Global Financial Strategies, 29 January 2014, www.credit-suisse.co m [20] CFA Program curriculum 2013: Equity and Fixed Income, Level 1 Vol. 5 p. 277 [21] Milton Friedman, Price Theory: A Provisional Text. Chicago, Aldine, 1962

[6] Campbell, Jeremy, R. Shiller, “Stock Prices, Earnings, and Expected Dividends.” Journal of Finance 43 (1998), 661-676.

[22] W.H. Beaver and M. Dale. What determines Price-Earnings Ratios? Financial Analysis Journal 34, no.4 (July-August), 1978.

[7] Anderson, Keith, and C. Brooks, “The Long-Term Price-Earnings Ratio.” Journal of Business Finance &

[23] J.W. Wilcox. The P/B-ROE Valuation Model. Financial Analysts Journal, 40, no. 1(January – February) 1984

6 Growth rate= ROE x (1- D/P ratio) where, Return on Equity(ROE)= 1/PE ratio (latest disclosed in December 2015)

[24] S.H. Penman. The Articulation of Price-Earnings Ratios and Market-to-book Ratios and the Evaluation of Growth, Journal of Accounting Research 34, no.2 (Spring) 1996

[25] S.E. O’Bryne. EVA and Market Value, Journal of Applied Corporate Finance (spring) 1996 [26] R. Banz. The relationship between Return and Market Value of Common Stocks, Journal of financial Economics, 9 (March), 1981 [27] W.S. Bauman and R.E. Miller. Investors Expectations and value Stocks versus Growth Stocks. Journal of Portfolio Management, (Spring) 1997 [28] S. Block. A Study of Financial Analysts: Practice and Theory, Financial Analysts Journal, vol. 55, no. 4: 86 – 95. 1999

(June), no. 3:663-682, 1977 [35] Frank K Reilly and Keith C Brown Portfolio and Investment Analysis, 10/e South-Western, 159-83, 2015 [36] John W Peavy III and David A Goodman, “The Significance of P/Es for Portfolio Returns”, The Journal of Portfolio Management, Vol.9 No.2 43-47 Winter, 1983. [37] Robert A Weigand and Robert Irons, “The Contraction and Expansion of Market P/E Ratio: The Fed Model Explained” The Journal of Investing, 17(Spring) 55-64, 2008.

[29] CFA Program Curriculum 2013: Equity and Fixed Income, Level 1 Vol. 5 pp. 262-66

[38] Robert A Weigand and Robert Irons, Market P/E Ratio, Earning Trends and Stock Return Forecasts”, The Journal of Portfolio Management, 33(Summer) 87-101, 2007,

[30] S. Skogsvik. Financial Statement Information, the Prediction of Book Return on Owner’s Equity and Market Efficiency: The Swedish Case, Journal of Business Finance and Accounting, 35(7) & (8), 795-817, Sept/Oct 2008

[39] Tim Laughran and J Wellman Imam, “New Evidence of the relation between Enterprise Multiple and Stock Return” Journal of Financial and Quantitative Analysis, 1629-50, 46 (2012)

[31] K. Skogvsik and S. Skogsvik. P/E-Ratios in Relative Valuation–A Mission Impossible? Investment Management and Financial Management, Vol. 5, Issue 4. 2008

[40] J. McWilliams. Prices, Earnings and P-E Ratios. Financial Analysts Journal, vol. 22, no. 3: 137. 1966

[32] E. Fama and K. French. Size and Book-to-Market Factors in Earnings and Returns, Journal of Finance, vol. 50, no. 1: 131–155.1995 [33] J. O’Shaughnessy. What Works on Wall Street. New York: McGraw – Hill. 2005 [34] S. Basu. The Investment Performance of the Common Stocks in Relation to their Price-Earnings Ratios: A Test of the Efficient Market Hypothesis. Journal of Finance, vol. 32

[41] P. Miller and E. Widmann. Price Performance Outlook for High & Low P/E Stocks. Stock & Bond Issue, Commercial & Financial Chronicle: 26 – 28. 1966 [42] S. Nicholson. Price Ratios in Relation to Investment Results. Financial Analysts Journal, vol. 24, no 1: 105 – 109.1968 [43] S. Imam, R. Barker and C. Clubb, “The use of Valuation Models by UK Investment Analysts”, The European Accounting Review, Vol.17, No.3, 503-35, 2008.

Abbreviations & Acronyms Abbreviations

Acronyms

Abbreviations

Acronyms

AVP

Average stock price

df

Degree of freedom

ANOVA

Analysis of Variances

F-crit

F- Critical value

BSX

Bangladesh Stock Exchange

GOB

Government of Bangladesh

BSEC

Bangladesh Securities & Exchange Comm

GBV

Group PE based value

CAR

Corporate Annual Report

GAAP

Generally Accepted Accounting Principles

CBV

Company PE based value

MS

Mean Square Sum

CSE

Chittagong Stock Exchange

NAV

Net Asset Value

DPS

Dividend Per Share

NBFI

Non Bank Financial Institution

DDM

Dividend Discount Model

PLCs

Public Limited Companies

DCF

Discounted Cash Flow

p-value

Correlation coefficient

DSE

Dhaka Stock Exchange

SS

Sum square deviation

Appendix Table1.1. Model1 prediction frequencies (%) (Non-parametric statistics) CBV=NAV

CBV=AVP

Bank

NBFI

Insurance

Bank

NBFI

Insurance

14.29

0

0

14.29

0

0

Fairly Valued 100±10% Overvalued 110≤150%

21.43

0

7.69

14.29

8.33

8.33

150≤ 200%

28.57

33.33

23.08

28.57

16.67

0

200%≤

28.57

66.67

69.23

42.86

66.67

92.31

7.14

0

0

0

8.33

0

Undervalued 50≤ 90%

Table1.2. Model 2 prediction frequencies (%) (Non-parametric statistics) GBV=NAV

GBV=AVP

Bank

NBFI

Insurance

Bank

NBFI

Insurance

7.14

0.00

0.00

0

8.33

0

110≤150%

28.57

8.33

7.69

28.57

25.00

0

150≤ 200%

35.71

33.33

15.38

7.14

16.67

15.38

200%≤

28.57

58.33

76.92

64.29

41.67

84.62

0

0

0

0

8.33

0

Fairly Valued 100±10% Overvalued

Undervalued 50≤ 90%

Table 2.1.Bank Parametric test: Reject Ho:AVP =CBV=GBV ANOVA: Single Factor Groups

Count

Sum

Average

21.35

13.00

256.40

19.72

Variance 93.61

37.85

13.00

560.94

43.15

725.65

44.93

13.00

536.40

41.26

271.41

Source of Variation

SS

Df

MS

F-statistic

P-value

F-crit

Between Groups

4403.84

2.00

2201.92

6.06

0.01

3.26

Within Groups

13088.01

36.00

363.56

Total

17491.84

38.00 Table2.2. NBFI Parametric test, Accept Ho: AVP =CBV=GBV

ANOVA: Single Factor Groups

Count

Sum

Average

Variance

19.75

11.00

484.65

44.06

1688.39

40.35

11.00

953.62

86.69

3028.49

25.41

11.00

818.41

74.40

3541.07

Source of Variation

SS

Df

MS

F-statistic

P-value

F-crit

Between Groups

10594.44

2.00

5297.22

1.92

0.16

3.32

Within Groups

82579.48

30.00

2752.65

Total

93173.93

32.00 Table2.3.INSURANCE Parametric test, Reject Ho: AVP =CBV=GBV

ANOVA: Single Factor Groups

Count

Sum

Average

Variance

16.35

12.00

254.70

21.23

127.87

50.51

12.00

688.06

57.34

418.75

50.20

12.00

671.84

55.99

386.52

Source of Variation

SS

Df

MS

F-statistic

P-value

F-crit

Between Groups

10057.52

2.00

5028.76

16.17

0.00

3.28

Within Groups

10264.51

33.00

311.05

Total

20322.03

35.00 Table 3.1. Bank Parametric test, Reject Ho: NAV =CBV=GBV

ANOVA: Single Factor Groups

Count

Sum

Average

Variance

35.23

13.00

284.69

21.90

31.67

37.85

13.00

560.94

43.15

725.65

44.93

13.00

536.40

41.26

271.41

Source of Variation

SS

Df

MS

F-statistic

P-value

F- crit

5.24

0.01

3.26

Between Groups

3596.84

2.00

1798.42

Within Groups

12344.65

36.00

342.91

Total

15941.49

38.00 Table 3.2. NBFI Parametric test, Reject Ho: NAV =CBV=GBV

ANOVA: Single Factor Groups

Count

Sum

Average

Variance

21.71

11.00

281.40

25.58

290.17

40.35

11.00

953.62

86.69

3028.49

25.41

11.00

818.41

74.40

3541.07

Source of Variation

SS

Df

MS

F-statistic

P-value

F-crit

Between Groups

22986.68

2.00

11493.34

5.03

0.01

3.32

Within Groups

68597.32

30.00

2286.58

Total

91584.00

32.00 Table 3.3.INSURANCE Parametric test, Reject Ho: NAV =CBV=GBV

ANOVA: Single Factor Groups

Count

Sum

Average

Variance

15.56

12.00

295.16

24.60

308.40

50.51

12.00

688.06

57.34

418.75

50.20

12.00

671.84

55.99

386.52

ANOVA Source of Variation

SS

Df

MS

F-statistic

P-value

F-crit

Between Groups

8236.73

2.00

4118.37

11.09

0.00

3.28

Within Groups

12250.29

33.00

371.22

Total

20487.02

35.00