Building A Better Value Strategy

We show you how to build a value strategy that returns 30% a year.

Like what you're seeing?
Try our FREE 30 day trial And test your strategies against historical data immediately

Written by Liam Flavelle on 11 August 2017

Summary

  • We refine our version of The 10 Minute Strategy.
  • The new model would have returned 30% a year for 17 years.
  • We start trading the strategy with real money.

This article concludes our review and refinement of our version of Lino Patti's 10 Minute Strategy (Dissecting the 10 Minute Strategy). Based on Benjamin Graham's ideas, Lino's strategy looks provides a way to quickly and easily screen beaten up stocks for gems that the market has mispriced.

While Lino then goes on to qualitatively assess each prospect, our strategy automated the process of selecting stocks and showed a 23% annual return over the last return. The problem? Like most value strategies over the last few years, our model has under-performed recently.

The New, Improved Version

As usual, we will use the InvestorsEdge.net platform to backtest our strategy. Further charts, statistics and historical position information can be accessed by clicking here or click on any of the charts in this article - to access the different versions just select the History option from the left menu and navigate between the models.

A backtest of our new strategy shows the following returns when backtested between 2000-17:

Best version

You can see that our annual returns for the strategy are 30%, with maximum drawdowns generally hovering around 20%:

Drawdowns

The value of our holdings dropped in line with the Russell 3000 in 2007-8, while the other major drop for the strategy was during the Euro crisis of 2011.

Market Cap Breakdown

The strategy would have invested predominantly in Small- and Micro-Cap companies, with valuations between US $50-2,000 million. This was highlighted as a weakness of the previous versions of the system which is still present in the new one. Restricting the stock universe to companies to US $125 million reduces returns to a still acceptable 24%.

How the strategy works

The new version of the strategy is remarkably similar to the original, with the universe of stocks that the strategy considers now being defined as:

  • All US stocks and depository receipts.
  • The last twelve month rolling sum of EPS greater than the annual reported EPS from 2 years previously (slight change from the original).
  • Net Current Asset Value greater than 0.

Over the years this has resulted in a pool of around 800-1200 companies that the strategy will consider opening positions in.

The strategy then ranks each stock in the universe by five factors (our original strategy used 7 factors) -

  • Change in Last Twelve Month's EPS from the previous quarter.
  • Number of consecutive dividend increases over the last 10 years.
  • Yield.
  • Price to Free Cash Flow ratio (lower is better).
  • Price to Sales (lower is better).

The main difference with this new version of our strategy and the original is in the price ratios used - Lino's used a stock's P/E and Price to Book ratios whereas the new one uses Price to Free Cash Flow and Sales, ratios which from my own research have performed better in the last 5-10 years.

Lastly, our model selects the top 10 stocks (rather than the 30 in the original version) with acceptable liquidity to open positions in.

Recent Returns

The major weakness of our previous system was in the poor recent performance of the strategy. Here are the rolling 3 year returns (telling us how profitable the strategy would have been if we had run it for each of the 15 3-year periods going back to 2000:

Period

CAGR

2000-2

52%

2001-3

83%

2002-4

71%

2003-5

75%

2004-6

48%

2005-7

38%

2006-8

25%

2007-9

24%

2008-10

29%

2009-11

37%

2010-12

33%

2011-13

44%

2012-14

50%

2013-15

40%

2014-16

40%

What this table shows us is that if we had started using the system at any time in the last 17 years it would have made money.

Would We Trade This Strategy?

For us to execute a trading strategy in real life, we need to have good reason to trust that it will continue to be profitable. To help us decide the likelihood of this, we have four-high level tests that our model should pass:

  • Investable - The strategy should be tradeable in real life and should scale. Our backtests include trading fees, so frictional costs are already taken into account. From a liquidity point of view, increasing the starting cash to $100k and $1m still resulted in good returns. Pass.
  • Intuitive - There should be logical risk- or behavioral-based reasons why the strategy works. We have attempted to create a strategy that buys cheap stocks which the market has wrongly priced, and have focused on factors that either identify quality stocks or mis-priced ones. Pass.
  • Persistent - The factors involved should work over long periods of time. Our tests have concentrated on the 2000-17 period. From an academic point of view, this would not be long enough to prove persistence. However, from our point of view at Investors Edge, we consider going through two major market shocks and a series of smaller downturns as enough to convince us the strategy works on a long-term basis. Pass.
  • Pervasive - Pervasiveness is more an ideal than a hard rule - our model should work across countries, regions and sectors. Testing across other individual countries shows returns of 10-20% a year with high Sharpe and Sortino ratios (which measure risk-adjusted returns). A key reason for this reduced return in other countries is almost certainly the value ratios (Price-to-Sales and Price-to-FCF) that we use - Price to Sales tends to work effectively in the US but not as well elsewhere. Having said that, a test of all 10 countries currently represented within the Investors Edge platform returned an annual gain of 30% a year with a high Sharpe ratio of 1.12. Pass.

Your Takeaway

One of the major reasons for the original strategy's recent under-performance was the value factors used - P/E and Price to Book don't seem to have the same predictive powers as they did in Benjamin Graham's time. This is probably because they are so popular and have become a victim of their own success.

Bringing in two ratios that have a better recent track record has dramatically improved our simulated results, and is a combination that I often use successfully in other models.

Would I trade this strategy? Yes! I will be putting real money behind it and providing monthly updates as I rebalance the portfolio.

Comments

You can read comments but to make comments you need to be logged in.

Comments
Like what you're seeing?
Try our FREE 30 day trial And test your strategies against historical data immediately