Professional Documents
Culture Documents
Hackers Guide To Investing
Hackers Guide To Investing
Contents
7 7 9 9 13
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 Notes on Data Sources . . . . . . . . . . . . . . . . . . . . . . 13 Specimen Accounts . . . . . . . . . . . . . . . . . . . . . . . . 14 Calculations . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 3.4.1 3.4.2 3.4.3 3.4.4 3.4.5 3.4.6 EBIT: Earnings before Interest and Tax . . . . . . . . 15 TACE: Tangible Capital Employed . . . . . . . . . . . 17 NCASH: Net Cash . . . . . . . . . . . . . . . . . . . . 18 EV: Enterprise Value . . . . . . . . . . . . . . . . . . . 19 ROC: Return On Capital . . . . . . . . . . . . . . . . . 19 UEY: Unleveraged Earnings Yield . . . . . . . . . . . . 19
3.5 3.6
4 Return on Capital
6 Shutdown 6.1
Colophon . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
List of Tables
3.1 3.2 3.3 3.4 Prot and Loss Account for BSY (BSkyB) for 12 m/e 30 June 2011 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14 P&L presented by Digital Look - reconciliation . . . . . . . . . 15 Balance Sheet - source: Digital Look . . . . . . . . . . . . . . 16 TACE calculation for BSY . . . . . . . . . . . . . . . . . . . . 17
LIST OF TABLES
Chapter 1 Booting up
1.1 Introduction
My email address is [email protected]. Please contact me if you have suggestions or corrections to make. This book is called Hackers Guide To Investing for the following reasons: I am a geek, and I wanted a way to write out and clarify my own thought processes, as much for the benet of myself as to others. The title of the book was inspired by John Walkers Hackers Diet [10], inventor of the hugely successful AutoCad. This book is very much an ongoing and evolving work, in true hacker style. Release early. Release often, as Eric S. Raymond would say.
CHAPTER 1. BOOTING UP
This chapter is an exploration of Greenblatts book You Can be a Stock Market Genius, also known as The Great Book with a Terrible Title. Hereinafter, I shall refer to it as the Genius book. Eventdriven sums up the general strategy of the book in the following post on Motley Fool [1]: Terrible name, but an amazing playbook on event-driven investing. He actually said in a later book that he regrets writing it as it was a little advanced for the complete novice investor, it just made a some hedge fund managers a lot of money. I couldnt agree more. The concept of investing in spin-os is actually a use of his magic formula investing strategy. Thats essentially a formula that assesses a company on 2 basic things: the companys historic return on capital; and the trailing p/e (strictly the EV/EBIT but P/E works). Assuming you pick companies with an above-average RoA and a below average p/e youll nd that the portfolio of stocks will signicantly beat the market portfolio (for the specics read the book The Little Book that Beats the Market). Theres ususally forced selling by large funds that cant or wont hold the new spin-o company (as they are just given shares in the new company), often because its too small so breaks size limits on their fund. Other times, investors dont even look at it (e.g. 9
10
CHAPTER 2. GREENBLATTS GENIUS BOOK CWC having too much debt) so sell o the new shares without even bothering to do any research. This is all selling pressure forcing the price down that is not related to fundamental investing reasons. This indiscrimate selling is know as a technical, caused by excess supply of shares in the market. This indiscrimate selling technical is what creates the low P/E situation. As the company doesnt yet have published data on data systems (like Bloomberg or Reuters), most people dont know what the true P/E or P/book is as most people dont systematically run through the ling documents of spin-os to check out proforma accounts, if the datas not on a screen somewhere, it gets ignored as it doesnt show up in any screens. This is why its so great - if you do a little bit of work (get the accounts from the listing document, usually on the spin-o or parent companys website), and work out the return on capital for the last year and discover its high, its potentially a high quality company (with the high RoC indicating that it has some sort of barriers to entry, or moat in buet terminology, like a brand or particular product edge). So if you nd a high RoA and low P/E spin-o, youre in a fortunate position as a) youre one of the few people thats done the work; and b) theres selling pressure for non-investing reasons that might be forcing the price down low enough to create a bargain opportunity; c) the data eventually feeds into the data systems like Reuters and institutional investors suddenly start liking the stock. This often creates decent short term gains (over 6-12 months). The idea is to hold them sell dodgier spin-os after the misvaluation corrects itself (anything from a few weeks to a a [sic] year). For the better businesses with a true moat, these become nice long term positions you bought at a bargain price once-upona-time. Also look for situation where mgmt are being awarded signicant stock options in the new business. This is great, as its one of those times where the animal spirits of free enterprise work great. Mgmt no longer shares in the equity of some slow-moving lumbering corporate parent but are now directly incentivised in stock of the spin-o - their actions are now creating value for both themselves and other shareholders like you. This is a great motivator to get management to deliver on a turnaround plan or deleveraging from high debt burdens or any other story that involves maximising the share price. This is whey they usu-
2.1. INTRODUCTION ally outperform in years 2 and 3, as managements plan of attack starts driving the fundamentals / protability of the spin-o. Ive probably averaged around 40% pa on average for of these positions Ive held. Theyre unlikely to be all your positions, but they put a great kicker in there for any portfolio.
11
12
This chapter focuses on Joel Greenblatts book, The Little Book That Beats The Market [5], hereinafter referred to as the Magic book, because it expounds Greenblatts Magic Formula Investing strategy. The basis for the book is to buy companies that are good and cheap. Good is measured by a high ROC (Return on Capital), and cheap is measured buy a high UEY (Unleveraged Earnings Yield). Greenblatt has stated that the Magic Formula is not suitable for use on utility or nancial stocks. Joel Greenblatts book, The Little Book That Beats The Market [5],contains information about his Magic Formula. Unfortunately, some of the nittygritty details were omitted, and this guide is an attempt to spellthem out. I provide MY INTERPRETATION of his calculations. I cannot vouch that they are, indeed, the ones he actually uses.
3.2
In this section, I present some things to watch out for from various data sources on the web. The watchword appears to be that nearly all sources have some problems somewhere along the lines, so you may nd it safest to look at the accounts yourself. 13
14 MSN Moneycentral:
EBIT can be wrong. For example, in BSY for y/e Jun 2011, EBIT was reported as being Prot Before Tax, having aggregated such items as nance costs. This is wrong.
3.3
Specimen Accounts
12 m/e 30 June Code REV
Table 3.1: Prot and Loss Account for BSY (BSkyB) for 2011 Description Amount Revenue 6597 Operating Expenses -5524 EBITDA 1405 Depreciation and Amortisation -332 Operating prot 1073 Share of joint ventures 34 Investment Income 9 Finance costs - 111 Prot on Investment Disposal 9 Prot before Tax 1014 Tax -256 Prot from continuing Operations 758 Prot from discontinued operations 52 Prot for year 810
OPN JV
We present extracts for BSY.L (BSkyB) accounts for 12 m/e 30 June 2011: Prot and Loss Account, as given in the nancial statements (table 3.1) Prot and Loss Account, as reported on the Digital Look website (table 3.2) Balance Sheet, as reported on Digital Look (table 3.3)
3.4. CALCULATIONS
15
Table 3.2: P&L presented by Digital Look - reconciliation Description Amount Notes Revenue 6597 Same Operating Prot 1073 Same Net Interest -102 = 9 (Inv Inc) -111 Prot Before Tax 1014 Same Discontinued Operating Prots After Tax 52 Same Prot for period 810 Same
3.4
Calculations
Our goal is to calculate ROC and UEY. In this section, I will show you how this is done.
3.4.1
EBIT is operating earnings before interest and taxes. Greenblatt [5, p 139] justies his use of EBIT as follows: EBIT (or earnings before interest and taxes) was used in place of reported earnings becasue companies operate with dierent levels of debt and diering tax rates. Using operating earnings before interest and taxes, or EBIT, allowed us to view and compare the operating earnings of dierent companies without the distortions arising from dierences in tax rates and debt levels. Greenblatt does not specify what to include and exclude from the calculation of EBIT, so I present a list of items to specically include and exclude. Include: Income from joint ventures Exclude: Extraordinary items Investment Income In our example from table 3.1, we obtain an EBIT of 1107 (1073 OPN + 34 JV)
16
Table 3.3: Balance Sheet - source: Digital Look Assets Non-Current Assets Property, Plant & Equipment 896.00 Intangible Assets 1406.00 Investment Properties n/a Investments 366.00 Other Financial Assets 275.00 Other Non-Current Assets 82.00 3025.00 Current Assets Inventories Trade & Other Receivables Cash at Bank & in Hand Current Asset Investments Other Current Assets
Other Assets Total Assets Liabilities Current Liabilities Borrowings Other Current Liabilities
3.4. CALCULATIONS
17
3.4.2
Greenblatt [5] denes TACE (Tangible Capital Employed) as Net Working Capital + Net Fixed Assets. He states that excess cash not needed to conduct the business was excluded from the calculation of Net Working Capital. He was not specic about how to calculate excess cash. A search that I conducted over the internet suggests that the cash required to run a business may vary from 3% to 10% of revenue, depending on the type of industry for the company. For simplicity sake, I have therefore stripped out all cash, and added back an assumed cash requirement of 5% of revenue. This seems the most logical approach. If you do that for BSY, you obtain a TACE of 1445 (see table 3.4 for a breakdown of the values).
Table 3.4: TACE calculation for BSY Tangible non-current assets 1619.00 (= 896 + 366 + 275 +82) Current Assets 2329.00 Strip out Cash -921.00 Working cash 330.00 5% * 6597 REV Current Liabilities -1912.00 TACE 1445.00
Using the calculations from MFI Diary [7], the following line items that you might see on Digital Look would seem to be included or excluded:
18
Line item Components of Net Fixed Assets Non-current assets: Property, plant, equip Intangible assets Investment properties Investments Other nancial assets Other non-nancial assets Components of Net Working Capital Current assets: Inventories Trade and other receivables Cash and bank and in hand Current asset investments Other current assets Current Liabilities: Borrowings Other Current Liabilities Notes:
inc exc exc - use 5% of revenues instead unk inc exc inc
1. inc means include in the calculation, exc means exclude, and unk means that it is unknown what the correct treatment should be 2. All non-current liabilites are excluded. 3. Stockopedia [8] says that working cash should be considered as part of the capital employed. A ratio of 3% - 10% is normal. A blended average of 5% might therefore be a suitable gure. 4. If Net Working Capital is negative, just use the value 0.
3.4.3
The Digital Look website seems like a fairly reliable source for a companys balance sheet. NCASH is net interest-bearing debt, and is used to calculate EV (Enterprise Value). In our BSY example, the gure to use is -1412 (calculation is done in table below.
3.4. CALCULATIONS Cash at bank/in hand 921.00 Current borrowings -8.00 Non-current borrowings -2325.00 NCASH -1412.00
19
3.4.4
EV is the market value of equity including preferred equity + net interestbearing debt. i.e. EV = M KT N CASH
(3.1)
where MKT is the market value of the equity. Net interest-bearing debt is the negative of NCASH. On 6 August 2011, MKT was 11762m, based on a share price of 671p and the number of shares in issue of 1752.84m. We there obtain EV = 13174 (11762 - -1412)
3.4.5
(3.2)
3.4.6
UEY is calculated as follows: U EY = EBIT /EV In our example, UEY = 8.4% (1107 / 13174). (3.3)
20
3.5
SharelockHolmes calculations
In this section, we look at how SharelockHolmes performs its calculations of ROC and UEY. We use Morrisons (MRW.L) for year-ended 31 January 2011. Some general information it provides: share price market capitalisation net debt interest prot before tax 284p 7365.5m -817.0 -30.0 874
Although preferred stock and minorites are not specied by SH (SharelockHolmes), they appear to be 0, according to the Digital Look website. Here are extracts from the balance sheet, as supplied by SharelockHolmes: Intangible Assets Tangible Assets Other Non-Current Non-Curr Assets Inventory Receivables Cash Other Current Current Assets Current Liabilities Net Current Assets Non-Curr Liabilities Retained Earnings Equity Funds 184.0 7,786.0 41.0 8,011.0 638.0 268.0 228.0 4.0 1,138.0 -2,086.0 -948.0 -1,643.0 2,463.0 5,420.0
SH makes the denitions for Unleveraged Earnings Yield: U EY = EBIT /EV (3.4)
where they specically cite UEY as being applicable to the Greenblatt calculation. They dene EBIT = (P BT + Interest) where, of course, PBT is Prot Befor Tax. They dene EV = M KT + N DEBT + P ref erred + M inorities (3.6) (3.5)
3.6. CRITIQUES
21
where MKT is Total Market Capitalisation, and NDEBT is Net Debt. They then give the calculation for ROC as follows: ROC = EBIT /T ACE where Tangible Capital Employed is dened as: T ACE = N W C + T A (3.8) (3.7)
where TA denotes Tangible Assets, and NWC is Net Working Capital, dened as: N W C = Inventory + Receivables + W CASH CL (3.9) where WCASH denotes Working Cash, and CL denotes Current Liabilities. Working with the above data, we obtain: EBIT = 874 + 30 = 904 EV = 7364.5 + 817 = 8181.5 T ACE = 7786 + 638 + 268 + 228 2086 = 6834 which leads to the nal results: U EY = 904/8181.5 = 11.05% ROC = 904/6834 = 13.2% (3.13) (3.14) (3.10) (3.11) (3.12)
These computed values for UEY and ROC agree exactly with SharelockHolmes.
3.6
Critiques
MFI is not without its critics and observations. In this section, we explore some of those criticims. In the thread Stock Contest Ends on magicformulainvesting Yahoo Group [9], Gray and Vogel, Analyzing Valuation Measures: A Performance HorseRace Over the Past 40 Years [4] has the abstract: We compare the investment performance of portfolios sorted on dierent valuation measures. EBITDA/TEV has historically been the best performing metric and outperforms many investor favorites such as price-to-earnings, free-cash-ow to total enterprise
22
CHAPTER 3. GREENBLATTS MAGIC BOOK value, and book-to-market. We also explore the investment potential of long-term valuation ratios, which replace one-year earnings with an average of long-term earnings. In contrast to prior empirical work, we nd that long-term ratios add little investment value over standard one-year valuation metrics.
An author [9] observed that (sales) growth is a stronger factor than profitability (ROA or ROE). Price (P/E) was the strongest factors in price, profitability and growth. Combining price and growth outperformed price and protability. Combining all three factors produced the best results. Another suggests: The sales growth or the earnings growth must give better result, you can also check if earnings growth is greater than sales growth, so you can detect if you sale more but with less prots You can also add a momentum signal: price greater than a quarter ago. So with the growth ratio you compare 2 years in the past. You can also check if the average eps expect for this year and next yearare greater than last eps. In another thread on Magic Formula Yahoo Group in 2012, a poster reported: I think the problem with the magic formula is that too many of the stocks in the top 50 have a lot of hair on them or are on the list when the past 12 months is a poor predictor of future returns. Stocks that have fallen in these categories have included: 1. For prot eductation [sic] stocks (like STRA, COCO, CECO, APOL) 2. home health care (AMED, AFAM, ) 3. Chinese RTOs (NEP, UTA etc) 4. Solar companies (PWER and GTAT) and 5. Bio-tech Pharmas () I suspect that is why JG has moved away from the individual stocks (even 25 to 30 gives a tremendous amount of volatility and is now pushing his big secret, which is much broader.
4.2
This section discusses ROITA. In Free Cash Flow Isnt Everything, Gannon [2] said that Warren Buett is interested in a companys return on net tangible assets, although it was unknown exactly how he calculates unleveraged return on tangible equity. He reckons that ROIT A = EBIT /N T A where NTA is Net Tangible Assets calculated as: N T A = (Receviables + Inventory + P P E ) (AP + AE ) (4.2) (4.1)
where PPE is net Property, Plant and Equipment (i.e. after depraciation), AP is Accounts Payable, and AE is Accrued Expenses. Exclude: cash and equivalents, prepayments, goodwill, other long term assets Basically, you exclude non-interest bearing current assets/liabilities on the basis that it does not cost the shareholders anything. Gannon used an average 23
24
of current and previous years balance sheet gures in the denominator of the equation. Based on gures for 2012 for WMT (Walmart), and using gures from the balance sheet for 2011 in order to create averages for NTA, we use ant EBIT of $26.56bn and and NTA of $100.62bn you obtain a ROITA of 26.40%, which is a pre-tax return. Assuming a tax rate of 35%, that would leave WMT with a normal return on unleveraged tangible equity of about 17%. Gannon says that value investors overfocus on free cashow. If a company can earn high ROITA, then it is better that it retain its earnings than distribute them as dividends. He oer the following thought experiment: lets say that an investor can earn a 10% return on his money, ignoring taxes. He thinks thats a stretch in 2012, though - the number is probably closer to 7%. Lets take it as 10%, though. If a dollar in WMTs pocket cannot compound more than 15% a year, then its likely to be a poor investment. There is a lack of a margin of safety. You know youself better than WMT, its past is probably better than its future, and net income probably overstates owner earnings. Generally, if you are investing in a business for its protable future growth, you want that future growth to be at least 150% of the annual growth you think you could provide using the same money.
4.3
Expected Performance
This is not a formula used in nance, but a concoction devised by me. So caveat emptor. Assuming that a share is fairly valued, then we might expect the total return over a year, t, to be given by t=y+e (4.3)
where y is the current yield on a share, and e is the growth in earnings. Technically we should be using forward yield, which is a multiplicative factor on the earnings growth, but I consider the assumption on additivity to be acceptable. The current yield is a known quantity. We assume that the earnings growth is a function of a capital return, r (for example the ROE of a company), and the dividend cover, c. C is the ratio of earnings to dividends, and is a measure of the amount of earnings retained by the company, and presumably ploughed back into expanding earnings. The earnings growth would be: e = r(1 1/c) (4.4)
25
(4.5)
So, if a company had a yield of 5%, a ROE of 20%, and a dividend cover of 2, then the expected return one year out would be: t = 5 + 20(1 1/c) = 15% (4.6)
Furthermore, we might also expect the share price to exhibit a reversion to the mean of its PE. Let us call that p. Then the expanded form for total return becomes: t=y+e+p (4.7) As with yield and earnings expansion, the total return would be multiplicative with p, rather than additive. We keep with the simplifying assumption, which allows us to see the individual components of expected return. In order to come up with a value for p, we assume that the current PE, p0 returns to a normal PR, pn , exponentially over a period n. Thus p = (pn /p0 )1/n 1 (4.8)
Thus, if a share is on a PE of 13, and a fair value might be 15, which is expected to revert over a period of 2.5 years, then we would calculate: p = (15/13)1/2.5 1 = 5.9% (4.9)
So we might expect an additional 5.9% uplift in share price performance in the rst year of the share price. The value of p0 is known, and pn should be chosen sensibly. We might use an average of the PE that the stock has traded for over the last decade, but a more sensible value would be 15 - the long-term average PE of the market. Greenblatt reckons that stocks revert to fair value over a 2-3 year period, so choosing a value of n = 2.5 seems sensible. Putting the above together: t = y + r(1 1/c) + ((pn /p0 )1/n 1) (4.10)
Showing the components separately gives us the opportunity to examine the likely contributions to share price performance. Below, I have tabulated some gures for companies trading in the UK, using n = 2.5. EPIC SP r BWNG 238 19.9 MCRO 438 55.2 SN. 592 20.8 c pn 2.18 9.75 1.51 12.9 4.37 15.0 p0 y% e% 8.3 5.5 10.8 10.0 6.6 18.6 12.3 1.9 16.0 p% t% 6.6 22.9 10.7 35.9 8.2 26.1
26
The above gures were taken at 23-May-2012. SP is the share price (in pence), and r, c, p0 and y were taken from Sharelock Holmes. The implied returns (MCRO at nearly 36%, for example), look too high, and are unlikely to be attained in practice. In chosing values for pn , I have used the lower of 15 (the long-term market average PE) and the mean PE that the company has traded for in the past. Subsequent Performance On 13-Apr-2013, the portfolio was revisited. The one-year performance of the shares listed above were: BWNG +84.0%, MCRO +30.4%, SN. +22.1%. All of these shares performed better than the All-Share index: ASX +13.1%. The results were excellent.
Chapter 5 Metrics
5.1 Introduction
5.2
FCFY (Free Cash Flow Yield) is dened [3] as FCFY = Free Cash Flow per Share/Current Market Price per share (5.1)
27
28
CHAPTER 5. METRICS
Chapter 6 Shutdown
6.1 Colophon
A The current version of this document was prepared using L TEX on Debian. A Previous versions of this document were prepared using L TEX on the following systems:
Slackware 13.37, the leetest Linux distro on the planet. Windows 7, using Cygwins latex packages
29
30
CHAPTER 6. SHUTDOWN
Glossary
FCFY Free Cash Flow Yield. 27 NCASH Net Cash. 18 ROITA Return on Invested Tangible Assets. 23
31
32
Glossary
Bibliography
[1] Eventdriven. C&W (CW.) Finals [online] Available <http://boards.fool.co.uk/hi-there-im-a-professional-investor-andive-12304660.aspx> [Accessed 28 August 2011]. at:
[2] Gannon, Geo. Free Cash Flow Isnt Everything [online] Available at: <http://www.gurufocus.com/news/175393/free-cash-ow-isnteverything> [Accessed 20 May 2012]. [3] Garcia, Alejandro. How to Calculate Free Cash Flow Yield. [online]. Available at: <http://www.earningsyield.net/87/how-to-calculate-free-cashow-yield/> [Accessed 02 June 2012]. [4] Gray, Wesley and Vogel, Jack. Analyzing Valuation Measures: A Performance Horse-Race Over the Past 40 Years Journalof Portfolio Management, Forthcoming. [5] Greenblatt, Joel. The Little Book That Beats the Market, John Wiley & Sons, Inc., New Jersey, 1st Edition, 2006. [6] Greenblatt, Joel. You Can be a Stock Market Genius, Simon & Schuster Inc., New York, 1st Edition, 1997. [7] MFI Diary. The Calculation. [online] Available at: <http://justadrone.blogspot.com/2011/03/calculation.html> [Accessed 20 August 2011]. [8] Stockopedia. How Does Magic Formula Investing Work?. [online] Available at: <http://www.stockopedia.co.uk/content/how-does-magicformula-investing-work-55911/> [Accessed 20 August 2011].
34
BIBLIOGRAPHY
[10] Walker, John. The Hackers Diet. [online] Available at: <http://www.fourmilab.ch/hackdiet/www/hackdiet.html> [Accessed 28 August 2011].
Index
Buett, 23 Free cashow, 24 Gannon, 23 Unleveraged tangible equity, 24
35