Work Breakdown Structure Table, Exercises of Project Management

Download Work Breakdown Structure Table and more Exercises Project Management in PDF only on Docsity! Perspectives E=7 4 Northill Capital Beta - A Surprisingly Complicated Greek Rick Potter, Partner Introduction Any investment professional or student of financial analysis will probably claim to have a good understand- ing of the concept of stock beta. However, conversations with a range of valuation professionals, both in the UK and elsewhere, appear to imply that few people really take the time to truly understand it. When questioned why they have calculated beta in a certain way the answer often comes back “because that is company policy.” As with many things in finance there is no one right way to arrive at a company’s beta. What I have sought to do in this brief discourse is to raise some of the questions that I think should be considered when trying to decide on an appropriate calculation methodology as opposed to simply firing up the nearest Bloomberg terminal, or your preferred equivalent, and reading off the number based on its default calculation parameters. Brief recap The Capital Asset Pricing Model (CAPM) and its derivatives are amongst the most widely used techniques for estimating the return, and hence value, of an investment in a risky asset. The basic CAPM model! has three parameters: i. the risk free rate (Rp) ii. the estimated market return (Rm) iii, Beta ameasure of the sensitivity of the asset's return to the returns of the market as a whole (its systematic risk). Figure 1 Expected Return Security Market Line Re Rr R,=R, + B(R,,- Ry) Beta Graphically, beta is the slope of the Security Market Line. Mathematically, it is the covariance of the Expected Return (R,) and Market Return (Rm) divided by the variance of the Market Return. Stock betas, at least for public companies, are typically estimated by regressing the returns of a stock against the returns of the market. So farso good. However, if we pause to consider the inputs Page 1 for amoment there are a number of questions that arise. In this paper I am not intending to be exhaustive, for example I park the question of “risk-free rate” selection, but simply to highlight some of the questions that should be thought about particularly when trying to derive a beta for use in valuing a privately held company. In particular: i. What “market” should we be using? ii, What time period and what frequency of measure- ment should be used? iii, How should we select a comparable peer group? iv. Should we use adjusted or unadjusted beta? Given our area of investing I have used companies in the asset management sector as examples. What market should we be using? Per the CAPM, the market measure should be the entire market of all risky assets, measured in a value-weighted index. In the real world no such index exists - to quote Eugene Fama? “poor proxies for the market portfolio of invested wealth is one of the key weaknesses of CAPM.” So we need to give the selection ofa suitable market proxy some consideration. In practice it is usually taken to be an appropriate stock market index. Since the correlation between stock indices is assumed to be at or close to unity, little thought is usually given to which particular index to use. However, consider figure 2 below. This shows the stock beta for Schroders plc derived by regressing the monthly returns on the stock against each of three equity indices over rolling 3-year periods, between March 1996 and March 2016. As can be seen, there are extended periods where the beta differs significantly depending on which index has been used. So index selection is an important consideration. Figure 2 25 2.0 05 0.0 Mar Sep Mar Sep Mar Sep Mar Sep May 1996 1998 2001 2003 2006 2008 2011 2013 2016 — FTSE 100 Source: Bloomberg == FTSE ALL Share == MCS! World Over what time period and with what frequency? As an example, consider the beta of the US manager Franklin Resources. The following table gives the beta, as Perspectives measured against the S&P 500, for various combinations of time period and frequency ending 31 May 2016. Pera ra) Prig Vee acon Period 5 years 1.469 1575 1.755 3 years 1371 1572 n/a 1 year 1.380 1.761 n/a Source: Bloomberg. As can be seen, the choice of period and frequency has a significant bearing on the result. Note that monthly results for 1-year and 3-year periods are not shown - it is normally accepted that at least 50 data points are required to produce a reasonable result. The problem of comparable peers Having decided how best to calculate beta, a common problem, at least when trying to determine the appropriate beta for a privately held company, is how to arrive at a beta for the target company given that there is no observable share price from which to calculate it. Many valuers will use the mean or median beta from what they consider to be an appropriate peer group. Again, using our own asset management sector as an example, let me illustrate the problem by trying to estimate the beta for say, Aberdeen Asset Management, by using a peer group. Assume that the peer group consists of listed European asset management companies. Aberdeen earns the majority of its income from “traditional” as opposed to “alternative” products so let's exclude those managers we might define as alternative managers. That leaves a peer group of eight managers. Amundi and Anima have not been listed long enough to provide a sufficiently long- term share price track record. If we exclude those two as well we end up with a peer group of six managers. The following table shows the beta for the peer group as at the end of May 2016 based on five years of monthly data. Schroders 1.524 Hendersons 1.728 Azimut Holdings 1.084 Jupiter 1.378 Charlemagne 0.844 Liontrust 0.837 Mean 1.233 Median 1.231 Source: Bloomberg. If we assume that the median is the appropriate measure then the peer group suggests that the beta for Aberdeen Asset Management should be 1.231. In fact Aberdeen’s beta over the same period was 1.823 - a difference of almost 50% which would give rise to a very different outcome in any discounted cash flow (DCF) based valuation. The problem is two-fold; firstly the number of listed asset management companies, even in the US, is really Page 2 E=7 4 Northill Capital too small from which to derive a meaningful peer group data set. Secondly, the idiosyncratic nature of active asset management companies means that there are usually very significant differences between them, hence their differing market sensitivities. In our sector these two issues raise the not unreasonable question as to whether or not CAPM is actually a suitable approach at all - but that is a subject for another day. Adjusted versus unadjusted beta The concept of adjusted beta was first proposed by Marshall Blume* of the University of Pennsylvania. Published in 1971, his paper considered “The Stationarity of Beta over Time.” He analysed stocks traded on the NYSE over six periods between 1926 and 1968 to see how the average beta of the stocks varied over time. He found that, for portfolios of stocks, their average beta regressed towards the market mean (i.e. a beta of 1) and that future (“adjusted”) beta can be estimated as 0.371 + 0.635 x historic (“unadjusted” or “raw”) beta. The concept of adjusted beta is widely used by valuation practitioners in their calculations. The table below shows both the unadjusted (raw) and adjusted betas for the same portfolio of asset managers (as measured against the FTSE 100 using five years of monthly data ending May 2016). re nee ee cedsre) Schroders 1.524 1.349 Hendersons 1.728 1.485 Azimut Holdings 1.084 1.056 Jupiter 1378 1.252 Charlemagne 0.844 0.896 Liontrust 0.837 0.891 Mean 1.233 1.155 Median 1.231 1.154 Source: Bloomberg However, in the same paper Blume also showed that this regression to the mean was not evident for small portfolios or single stocks. Consider figure 3 below. This shows how the beta of four listed US asset management companies has varied over time. Figure 3 25 2.0 1.5 ehh 1.0 0.5 0.0 Apr Oct Apr Oct Apr Oct Apr Oct Apr Oct 1993 1995 1998 2000 2003 2005 2008 2010 2013 2015 = Alliance Bemstein — Franklin Resources Mean — Eaton Vance — Legg Mason Source: Bloomberg.