Monday, August 29, 2016

Maximizing Returns with Index Investing

Alpha, or excessive returns greater than the market average, is extremely hard to obtain over long periods of time.  If you are getting suckered into marketing tactics that market 20% returns over long periods of time, you need to READ THIS IMMEDIATELY.

That is why I look to hold a large portion of my portfolio in index ETFs that simply provide the returns of the market.

That said, I have done some research with AmiBroker recently and have identified how I am going to approach the market.  Here was/is my objective as it relates to the indexing portion of my portfolio:

Objective: To beat the returns of the S&P500 (SPY) by combing index funds that have provided returns greater than those obtained by simply holding SPY.

It is important to note a subtle point here.  I am not trying beat the market.  I accept that beating the market is very hard and I am competing against the best of the best professional money managers (who have trouble with this too).

What I am trying to do it maximize my returns - get the highest return I can without taking on too much extra risk.


I made a great investment a few years ago and that was by buying a copy of AmiBroker.   AmiBroker allows investors to set up trading/investing systems and backtest them to see how they would have done in the past.

Of course, we have no idea what the future holds and just because a system I put together in AmiBroker looked good in the past does not mean it is going to do well in the future.

However, if done properly it can increase the chances of seeing similar behaviour in the systems.

Using AmiBroker, here is the approach I took to develop the index investment system I currently am using:

  • Use consistent dates of the backtest to generate the appropriate baselines.  I used 2007-01-01 to 2016-08-26 as my standard dates for the test.
  • All tests will be compared to a simple buy and hold investment in SPY.  In other words, I buy SPY on 2007-01-01 and see what the value is on 2016-08-26.  If any systems generated do not beat that benchmark, then why bother?
  • The index products I will use are general market index funds, that track different parts of the market.  The sky is the limit in terms of which products I could have test with, however I chose the following four as they were the ones I would be comfortable holding for years to come:
    • SPY:  Pretty self-explanatory.  Tracks the S&P500 as a market-weighted index ETF.
    • RSP: Similar to SPY, except that it tracks the S&P500 with an equal-weighted approach.  In other words, every stock in the S&P500 is held as an equal-weight holding.  The SPY is market-weighted meaning that stocks like Apple and Google tend to skew the returns based on how those stocks do.  RSP simply buys the whole market in an equal-weight.  There are caveats to RSP, however this article is what got me thinking about including it in my portfolio.
    • MDY: MDY tracks a market-cap-weighted index of midcap US companies.  Mid-caps have performed better than large-caps historically, so I wanted to see if adding these to my portfolio would help get better returns.
    • QQQ:  The Q's includes 100 of the largest domestic and international nonfinancial companies listed on the Nasdaq Stock Market based on market capitalization.  In other words it tracks a lot of technology stocks.  The Q's has done better than SPY for years, so again I wanted to see if I could add returns by diversifying using this ETF as well.  In fact, it has done so well, it is the only ETF that I included in all test portfolios.
  • Standard testing settings will be applied, including commission costs of $0.005/share, buying at the open, and starting with an equity position of $75,000.  The equity is dividend into equal parts of each holding.
  • So as to not get sucked in by better returns, I am going to use the Sharpe Ratio, Maximum System Drawdown (MDD), and Recovery Factors as my key performance indicators.  Everything else being equal, the portfolio with the best combination of these three will be called the "winner".
Hope that all makes sense.  Let's get to the results.

Results: SPY, MDY, RSP, and QQQ Buy and Hold Performance

To start, I tested simply holding each one on of these ETF's separately to see how they would do compared to each other, as well as set a benchmark for a buy and hold of the SPY.

This chart provides a summary of a simple buy and hold of each of these ETFs:

The chart is self-explanatory, however just to be clear I will explain the results using the SPY as an example.  The portfolio started on Jan 3, 2007 with $75,000 and ended on July 26, 2016 with $139,514 for a 86% gain or 6.64% annual return.

That is a pretty good run and solid returns any investor who chooses the easy task of just buying and holding without doing anything stupid (like trading on emotions) could easy achieve.

As you can now see from the chart, an investor could have beat the SPY simply buy buying any of the other 3 ETFs; the MDY,  RSP, and QQQ's all beat the SPY over the same period.  In fact, a simple investment in the Q's destroyed an investment in SPY and earned 11.77% per year for a total return of 193%.

The task now is to see if I can beat these returns on a risk adjusted basis.

Results - Basket Buy and Hold Analysis

Now that we can see how just buying the SPY would have done for us, and that even better returns could have been had by investing in either the mid-cap portion of the S&P (MDY), the equal-weighted representation of the S&P 500 (RSP), or the technology heavy Q's (QQQ), we can now look at how combining them would have done.

The reason that an investor might want to hold more than one of these ETFs is due to diversification.  Holding just one of these at a time means we are putting all our eggs in that one index basket.  Since I have chosen four indexes that trade slightly different markets, the theory is that we can get better returns even with an even less concentrated portfolio.

For example, a portfolio of SPY, MDY, and QQQ means that an investor gets the returns of the S&P 500, the S&P mid-cap market, and the technology (non-financial) portion of the market.  Since indexing is all about diversification while capturing the returns of the market, why not see if even broader diversification can be had while getting better returns.

Using the same methodology as above, here is how the performance stacked up, in four combinations of portfolios:

You can read this chart the same way as the last one.  For example, a portfolio that buys an equal dollar amount of SPY, MDY, RSP, and QQQ would have turned $75,000 into $168,234.  That is a 124% return or 8.73% CAR.

It is interesting that this initial portfolio, which is super-diversified does better than a simple SPY portfolio.  It provides an extra $28,710 of return.

All four portfolios do progressively better, ending with the best portfolio being a combination of MDY, RSP, and QQQ.  This portfolio, which holds the mid-cap market, an equal-weighted S&P500 portfolio, and the non-financial portfolio of the market via the Q's, returns 136% or 9.31% per year.

That is nicely above the SPY's 6.64%.  But it gets better when I had a look at some other key metrics like the Sharpe Ratio, MDD, and Recovery Factor.  Have a look at how these portfolios do - and keep in mind that a simple SPY portfolio has a MDD = -55.18% and a Recovery Factor of 1.40.
Sharpe Ratio: The sharpe ratio is a tool to examine the performance of an investment by adjusting for its risk. The ratio measures the excess return of a trading strategy.  The higher the better.

Recovery Factor: The ratio of net profit and maximum system drawdown. The higher the score the better.

As you can see the "MDY RSP QQQ" portfolio does well compared to the other three portfolios with a higher sharpe ratio, similar MDD (although it is not the lowest), and strong Recovery Factor.  Based on these results, it would have been the portfolio to hold.

To be more specific, an investor gets similar drawdowns, but better recovery and a higher sharpe ratio to go along with the higher returns.  I am going to be using this combination of index funds in my portfolio.


Although I am not very good at predicting the future, I feel that this analysis has provided me with a solid foundation that could allow me to earn returns that are better than simply an investment in SPY.

An investment in three index funds - MDY, RSP, and QQQ - has provided better returns with a better risk profile.  The drawdowns are similar, but the returns we get in return are good making the 

There are going to be times when the market goes down and this portfolio will go down with it.  However, I will be holding this portfolio for in excess of 10-15 years so will be able to weather whatever the market throws at me.

And although nothing has beaten an investment in just QQQ,  the added diversification of this portfolio will allow me to sleep better at night.

No comments:

Post a Comment