Investing in mid-large caps companies with high yields and returns on assets has historically delivered superior returns.
Backtests show average returns of 15.2% annually since 2000, including an average dividend yield of 3.2%.
The strategy has returned 5.4% in the 4 months since its we started tracking it in October.
Back in October, Fred Piard produced an interesting piece of research here detailing a very simple income strategy that invests stocks with high yields and returns on assets.
Since imitation is the sincerest form of flattery I decided to confirm the results of his findings using the InvestorsEdge.net platform. Once I’d added some of my own tweaks, I liked what I saw and have been following the strategy with a theoretical pot of money ever since.
Backtested Strategy Returns
Detailed risk/return information and position data together with the a record of how I refined Fred’s original model can be found on the InvestorsEdge platform here – below are the simulated results of trading the strategy for the last 18 years:
You can see that the strategy shows an average annual return of 15.2%, compared to the S&P 500’s 3.2% performance. The Sharpe ratio, which together with the Sortino ratio measure our returns based on the risk we assume running this strategy, is 0.84 compared to the S&P 500’s 0.26.
Risk and return are very much a subjective decision – 0.84 for a large cap strategy is just about acceptable for me.
The strategy coped well with market downturns, with only 10 drops greater than 10% over 18 years, and only two greater than 20%. The drawdowns chart shows the peak to trough drops in value of our model over the tested period, and also the time each drop took to return to profitability:
The drawdowns chart is the first one that I tend to look at in a backtest – having great returns is one thing, but if the graph looks like a roller coaster ride at Disneyland then it’s a no go for me.
How the Strategy Works
Our strategy is extremely simple – each month all US stocks and ADRs with a market capitalization greater than $5bn that yield between 3% and 10% are selected.
This list of stocks is then sorted based on their Return on Assets, with larger values ranking higher. We then invest in the top 20 stocks, limiting our total exposure to any one sector to 25%. Old positions are exited unless they are part of the new list of stocks.
Why The Strategy Works
Put simply – quality! This strategy strives to identify quality companies using two criteria, yield and efficiency.
The first quality marker for a company is the fact that it pays a dividend. This is by no means full proof, but my research has consistently found that companies that return cash to shareholders tend have more consistent cashflows and profits and also tend to exist at a later and less volatile stage in their development than their peers.
As a downside this also means that dividend-paying companies have actively chosen not to invest that money into their own business. This is where the second factor comes into play.
Our second factor selects companies with higher return on assets when compared to their peers. ROA is calculated as :
This infers that companies with higher returns are more efficient – they wring more profits out of each dollar of assets that has been invested in the company. This efficiency could represent a wide economic moat that fends off market entrants, a highly skilled management, or even technological advantages, but they all have one thing in common – they can represent a more desirable and safer place for an investor to place his or her money when compared to less efficient competitors.
The more cynical amongst us (and that includes me) will be thinking that backtests always look good – hindsight can be a wonderful thing. Once I have refined a strategy, I will run it with a theoretical pot of money to test that it actually works in real life.
I have have been tracking the results of an imaginary $10,000 investment in this strategy since October, and you can see the results on the InvestorsEdge site here – the highlights are below:
The above results show the S&P 500 beating our strategy, but need to be taken in context of the recent downturn in stocks:As you can see from the above chart, up until a 2nd March our strategy was beating its benchmark, and I have no reason not to believe that the strategy will quickly return to out-performing the index. Indeed, keeping up with a market that is generally punishing high yield stocks (the average yield in the current portfolio 20-stock is 4.4%) is pretty remarkable in and of itself.
The strategy currently holds the following companies:
A key risk that I always examine with mechanical investing strategies is that the data phenomenon that I am exploiting will simply stop working. To combat this, I look to see if a strategy intuitively makes sense – our model invests in companies that make the most efficient use of their assets and also pay us an income, which to me makes sense.
Entering positions in stocks can be easy, but getting out again can be considerably more costly in a downturn. The companies selected for our strategy have a market cap greater than $5 billion, which reduces many of the liquidity risks associated with exiting positions.
The macroeconomic climate can have an impact on any strategy and this one was not immune to the 2008 market crash. As seen from the drawdowns chart above the strategy would have bounced back fairly quickly to pre-crash levels by 2010.
A specific macroeconomic risk is the market view of interest rate increases. With traditional high yield stocks being punished over the last 6 months, our strategy has demonstrated a good degree of resilience.
Data anomalies are also a potential risk. Sometimes we data mine and exploit an anomaly that doesn’t repeat itself. We tend to look at the fact the strategy makes sense and has fewer working parts to see if this is the case and we test in different time frames to identify data mining. If we are happy that the strategy is robust we will then run it using a theoretical pot of money to see if it performs in line with the backtests..
Having tracked this strategy for 6 months and seeing how it has reacted to a small downturn I believe that it is a valid candidate for putting some real money to work, which is exactly what I will be doing from now on.