WHAT WORKS ON WALL STREET – James P O’Shaughnessy

 

Amazon link

 

In a nutshell

This document is just a brief summary of the main points in the above book.  In a nutshell, it’s a book advocating the disciplined application of mechanical investment strategies (i.e. objective, unambiguous, unemotional strategies) and providing vast amounts of evidence collected since 1950 to demonstrate that they work!

 

Post nutshell Pre-Amble

Most investment strategies are mediocre and the majority, particularly those most appealing to investors over the short term, fail to beat the simple strategy of indexing to the market (e.g. FTSE 100 Trackers)

Rather than supporting the academic theory that stock prices fluctuate in a “random walk”, it can be demonstrated that the stockmarket methodically rewards certain strategies whilst punishing others.

The only way to beat the market over the long term is to consistently use sensible investment strategies.  Discipline is the key.  Over 80% of mutual funds fail to beat the index because their managers lack the discipline to stick with one strategy through thick and thin.  This lack of discipline devastates long term performance.

 

Main points

·         Most “small cap” strategies owe their superior returns to microcap companies but these stocks are too small or too illiquid for anyone to sensibly invest in

·         Buying stocks with low PE is most profitable when sticking to large, better known stocks

·         Price to Sales ratio (PSR) is the best “value” ratio for buying market-beating stocks

·         Last year’s biggest losers (i.e. lowest 12 month relative strength) are the worst stocks you can buy

·         Last year’s earnings gains alone are worthless when determining whether a stock is a good investment

·         Using several criteria dramatically improves long-term performance

·         Relative strength (i.e. change in share price over a year) is the only growth variable that consistently beats the market

·         Buying current “darlings” with the highest PE ratios is one of the worst things you can do

·         A strategy’s risk (e.g. Sharpe Factor (see below)) is one of the most important elements to consider

·         Using growth AND value strategies combined is the best way to improve performance

·         Low PSR is the best single factor strategy in terms of absolute returns but risk can be reduced by combining it with other factors

 

 

Definitions

Active Investor	Someone that selects stocks they believe are superior to others.  They are typically guided by one of two styles – value or growth
Passive Investor	An investor that uses index tracking
Value Investor	An active investor that looks for stocks with market value below true or liquidating value.  They use factors like PE ratio, price to sales ratio, price to book ratio, etc. to identify stocks selling below their intrinsic value or the value of their underlying assets
Growth Investor	An active investor that looks for stocks that have higher than average increases in sales and earnings with expectation of more of the same.  A classic growth stock’s earnings just keep on getting better and better.  Growth investors believe in a company’s potential and that a stock’s price will rise with earnings
All Stocks	A database of stocks from which the micro capitalisation companies have been removed [Basically seems to equate to top 2000 companies listed in UK]
Large Stocks	The subset of “All Stocks” that have a market capitalisation greater than the market average [Basically seems to equate to the FTSE 350]
Sharpe Factor	A formula to quantify the risk of a particular investment strategy (based on historic returns).  The Sharpe Factor is calculate as (Annualised Rate of Return – Risk Free Rate of Interest) / Standard Deviation of Returns (as a decimal 0<sd<1) For example, let’s say that the risk free rate of interest is 3%.  If Strategy 1 returns an annualised rate of return of 20% but is very volatile so has std dev = 0.33 (i.e. 33%) then the Sharpe Factor is (20-3)/0.33 = 51.  If Strategy 2 returns 16% but is less volatile with std dev = 0.20 (i.e. 20%) then the Sharpe Factor is (16-3)0.2 = 65.  Strategy 2 is therefore said to have the better “risk adjusted rate of return” and is generally considered to be preferable to Strategy 1 (although this clearly depends on the investor’s risk threshold)

 

Rules for defining a Mechanical Strategy

A Mechanical Strategy needs to keep emotion out.  It must be:-

·         Explicit – there must be no ambiguity in the rules and they must not be open to private or unique interpretation

·         Public – must be stated explicitly and publicly so anyone can reproduce the result.  The strategy must make sense and must not be derived from data

·         Reliable – someone testing the same selection criteria against the same data must get the same results and the results should be consistent over time (i.e. a strategy should not owe its good performance to just a couple of key years)

 

BackTesting

Backtesting should use a reliable database which avoids:-

·         Data Mining – looking for relationships between data in order to derive rules – the rules should be translatable from one set of data to another

·         Limited Time Periods – only having data for a short period – especially a short “bull” period

·         MicroCap Stocks – these small companies skew results and are unrealistic as they are almost impossible to trade in without huge spreads

·         Survivorship Bias – only including stocks that exist now.  This overstates the results as it ignores all the companies that have failed (and which could still have been selected by the strategy if it had been used for real)

·         Look Ahead Bias – basing decisions on fundamental information that wouldn’t have been available at the time (e.g. using earnings data in January when it may only have been reported in March)

[The exercises through the book use a 50 stock portfolio with rebalancing each year – shares purchased in equal money amounts.  Transaction costs were ignored.]

 

Summary

·         Always use strategies

·         Only use strategies proven to work over the long term

·         Invest consistently

·         Consider your risk tolerance

·         Plan your path and stick to it – don’t jump in and out and don’t look for companies that just have a great story

·         Never use the riskiest strategies (they’ll sap your will – use those strategies with best risk adjusted returns)

·         Consider how to balance the different levels of risk in strategies in a way that you’re comfortable with in terms of your risk tolerance

·         Use more than one strategy in parallel (always have some growth and some value as diversification)

 

Summary of Returns for the different Strategies

There are pages of tables in the book and performance results of at least 50 different strategies over 45 years are provided.  The table below is a selected extract of the tables and the strategies are listed in descending order of Sharpe Factor (i.e. strategy with best risk adjusted rate of return is at top)

Strategy	Annualised Rate of Return	Std Dev of returns over 45 years	Sharpe Factor
United cornerstone strategies (see below)	17.1	19.5%	66
Cornerstone Value strategy (see below)	15.06	16.47%	62
PSR <1, All stocks, high rel strength	18.62	25.64%	61
Cornerstone Growth strategy (see below), All stocks	18.52	25.41%	61
+ve Earnings yield over last 5 years, All stocks, high  rel. strength	18.52	24.48%	61
Price to Book <1, All stocks, High rel strength	17.95	23.36%	60
Cornerstone Growth strategy (see below), Large stocks	14.71	17.5%	58
….	….	….	….
All stocks (i.e. the index)	12.91	19.46%	44
….	….	….	….
Large stocks (i.e. the index equiv to FTSE 350)	11.54	15.72%	41
….	….	….	….
High PE, All stocks	8.97	26.75%	23
….	….	….	….
Risk free interest	5.63	2.71%	0
….	….	….	….
Low 1yr rel strength	2.62	26.33	-1

 

The United Cornerstone Strategy is as follows:-

·         Invest 50% of your capital in the Cornerstone Growth strategy

·         Invest 50% of your capital in the Cornerstone Value strategy

This strategy beats the all stocks index tracker in 80% of the 46 years tested (the years 1951-1996), in 34 out of the 40 rolling 5-year periods and in 100% of the rolling 10-year periods.

 

The Cornerstone Growth Strategy is as follows:-

·         Use the All Stocks list of stocks

·         Use only stocks that have eps > last year’s eps

·         From this subset, use only stocks that have PSR <1.5

·         Rank this subset by high relative strength and purchase those at the top of the list (O’Shaughnessy’s strategies are all based on 50 stock portfolios – i.e. he’d buy the top 50 from this list in equal money amount (and hold for a year then do the same again))

This strategy beats the all stocks index tracker in 78% of the 46 years tested (the years 1951-1996), in 33 out of the 40 rolling 5-year periods and in 94% of the rolling 10-year periods.

 

The Cornerstone Value Strategy is as follows:-

·         Use the Large Stocks list of stocks

·         Use only stocks that have more common shares o/s than the average

·         From this subset, use only stocks that have cashflow per share exceeding the average

·         From this subset, use only stocks that have sales (or turnover) at least 1.5 times the average

·         Exclude all Utilities stocks from the subset (as they tend to dominate these criteria
(O’Shaughnessy calls the stocks that satisfy the above constraints as “Market Leading Stocks” – it generally is around 6% of the stocks in the market as a whole)

·         Rank this subset by high dividend yield and purchase those at the top of the list (O’Shaughnessy’s strategies are all based on 50 stock portfolios – i.e. he’d buy the top 50 from this list in equal money amount (and hold for a year then do the same again))

This strategy beats the all stocks index tracker in 64% of the 46 years tested (the years 1951-1996), in 35 out of the 40 rolling 5-year periods and in 97% of the rolling 10-year periods.