ETF Complete Strategy Insights: Measuring A Model (Part 4)

James Kimball | May 19, 2019

The ETF Complete model closed the week down -2.8% compared to the SPY which closed down -0.8%.

After an initial gap lower on Monday off the heels of negative news and developments in the on-going trade negotiations with China, the SPY managed to put in green days the rest of the week, closing just off of the levels from last Friday.

Stay tuned for the daily updates and log into the website to see holdings and additional performance data.


This Week's Strategy Lesson: Measuring A Model (Part 4)


This week we wrap up this series looking at different risk and performance metrics. Building on the concepts of standard deviation introduced last week, we are going to get into the details of how the Sharpe and Sortino Ratios are calculated and what they tell us about risk-adjusted returns.

Risk-Adjusted Returns

When you look at just performance, you could be comparing instruments or trading strategies that are vastly different in terms of the relative risks being taken or its appropriateness for different investing contexts.

It’s possible that a model that was only up 5% in a given period was actually taking considerably more risk than one that was up 10% (or vice versa). Risk here is measured using the standard deviation calculation. Standard deviation measures the volatility of a stock or trading model. The higher the volatility the higher the chance that the stock or trading model will have performance different from its historical average (positive or negative).

It should be noted here that standard deviation relies on a normal distribution (see last week’s article) and stocks are not exactly normally distributed. Stocks tend to have more frequent extreme outcomes (outsized daily moves or gaps) relative to a normal distribution. Because of this, these measures of risk become more “guidelines” or “best guesses” about the risks we will see in the future.

Sharpe & Sortino Ratios

The Sharpe ratio is a common ratio used to compare the annualized returns of a model or a fund to its volatility (standard deviation).

The Sortino ratio is similar to the Sharpe ratio with the main difference being that instead of using the standard deviation of the whole sample, it only focuses on the standard deviation of the negative return months. The reason for this is that, in theory, as an investor we are mainly concerned about large negative moves and we might like large positive moves.

For some strategies, the two ratios might be very similar, signaling that the there is little difference between the positive and negative volatility. However, if the volatility is skewed in one direction, the Sortino ratio will pick this up (and in fact we see below that the ratio of the upside volatility compared to downside volatility significantly favors the ETF models over the S&P 500).


The data table above shows all the components we need to calculate the Sharpe and Sortino ratios. The Sharpe ratio is basically the average annual return divided by the average annual standard deviation (the formal calculation includes the “risk free rate” but we can ignore that for the periods we are looking at). The Sortino ratio is the average annual return divided by the average annual negative return standard deviations.

A Sharpe ratio around 1 or higher is considered good as far as balancing the risk and return profile of an investment strategy (and for the most part the higher the better). The Sector Moderate, Aggressive, and ETF Complete models are all around 1 while the Country and Sector Conservative model lags a little behind but still well over the 0.49 score in the SPY benchmark.

These numbers can be compared to each other. A score of 1 means you are getting twice the return per unit of risk compared to a score of 0.5. So it’s a form of an efficiency rating. The Sector Moderate and Sector Aggressive have very similar Sharpe ratios of 0.96 and 0.89. The Aggressive model is significantly more volatile than the Moderate model, but they both use their different levels of volatility equally efficiently.

The Sortino ratios for all of the ETF models are about 50-80% higher than their Sharpe ratios compared to the SPY Sortino ratio only being about 25% higher than its Sharpe ratio indicating that downside volatility in the ETF models is less than the SPY benchmark.

We hope that this series has provided useful information for evaluating the risk profiles of not only our ETF models, but any trading system or fund. Many of these metrics are industry standard and can be used to make better more informed investing decisions.

After a solid close to last week on the heels of a strong Jobs report and GDP report earlier in the month, markets dropped over the continual drip of negative news reports about the trade negotiations with China, though they close well off their lowest levels.