Summary

  • In due diligence, we often evaluate summary statistics like annualized return, volatility, alpha, beta, up-capture, and down-capture. 
  • These statistics can unify years of returns into a single number.
  • While this can be convenient for comparing different strategies, it fails to provide adequate insight into the actual week-to-week experiences an investor will face.
  • We highlight how even in the best performing years, a simple tactical strategy can exhibit nauseating relative performance on a week-to-week and month-to-month basis.

 

A few months ago we wrote a commentary on the concept that investors who look at their portfolio on a daily basis are more likely to think the market is riskier than those that look at their portfolio on a yearly basis.

This math works out because while the day-to-day probability of the market being up or down is near 50-50, the slightly increased probability of up days compounds over time to give year-to-year returns a higher likelihood of being positive.

For those of us who manage money professionally, we don’t have the luxury of not watching the market daily.

This is a problem for advisors building portfolios because they build these portfolios based on the evaluation of long-term summary statistics like annualized return, volatility, alpha, beta, up-capture, and down-capture.

Together they can paint a picture of full market cycle performance.

What they don’t tell you, however, is what the investor experience is on a day-to-day basis.

We evaluate and select using multi-year summary statistics but live and experience the portfolios on a day-to-day basis.

It can be difficult for the day-to-day emotions to not overwhelm the long-term outlook, even if the desired long-term results are still being realized.

Let’s consider a very simple tactical S&P 500 model.  When the S&P 500 is above its 200-day moving average, we invest.  When it is below, we divest and hold 0% returning cash.

The highlight reel for this strategy is made up of years like 2000, 2001 and 2008.

In 2001, the SPDR S&P 500 ETF “SPY” was down -11.8% while the tactical model returned 0.0%.  In 2002, SPY was down -21.6%; the tactical model was only down -5.8%.  Finally, in 2008, SPY was down -36.8% and the model was down -3.9%.

In retrospect, those numbers sound fantastic.  Who wouldn’t prefer -3.9% to -36.8%?

But they are summary statistics.  They compress a full year into a single number.  They don’t tell you what your actual day-to-day experience in 2008 was like.  They don’t tell you about the gyrations that occured in each and every one of the 52 weeks.

Let’s actually look at the the relative performance of the tactical strategy versus SPY in these periods.

Below we plot the year-to-date relative performance between the two strategies in basis points.

2000-2001

We can see that while the tactical model ultimately out-performed by 1100 basis points (bp) in 2001, there were many multi-week periods that may have caused us to pull our hair out.  Early on, the model gave up 400bp to SPY – and actually began to under-perform year-to-date.

Just a few months later, the model gave up 1250bp to go back to near parity with SPY.  In the 4th quarter, the tactical approach gave up 1400bp in relative performance.

The relative performance is a roller-coaster, and those are stomach-dropping periods of rapid under-performance.  To really test your mettle, imagine if you had invested in the tactical strategy at one of the relative peaks rather than at the beginning of the year.

2002 was much the same story.

2001-2002

Again, the tactical model ultimately out-performed, but there were many periods where it gave up hundreds, if not thousands, of basis points in relative performance.

By April, after giving back 750bp – and trailing the market by 500bp – would we have just walked away?

And then there’s 2008.

2008-2009

While Q1 was strong, all the relative out-performance was given back in Q2.  How many clients would be happy to underperform the market by 1100 basis points in a single quarter?

How good would it feel to underperform the market by 1309bp in a single week?  That happened from 11/21/2008 to 11/28/2008.

We bring this up only to highlight the complexities of choosing a strategy based on summary statistics but living with it on week-to-week basis.

A tactically risk-managed strategy, by definition, will go through periods where it looks very different than the benchmark.  That’s how it seeks to protect against losses: by not looking like the benchmark.

Even if the long-term call is correct, there are weeks, months, and quarters when not looking like the benchmark doesn’t feel very good.  This holds particularly true for bear market rallies, where short-covering can lead to incredibly swift and violent “melt-ups.”  These can be a painful time to be on the sidelines.

At the end of the day, no investor experiences the long-term summary statistics in isolation.  Instead, they experience a series of weeks and months, many of which can be emotionally trying and turbulent.  Over time, the results of these periods accumulate to look, on average, hopefully something like the summary statistics.

So we must remember: summary statistics are just long-term averages, and we know that “average” is not “annual” and it is certainly not what you should expect on a week-to-week basis.

Corey is co-founder and Chief Investment Officer of Newfound Research, a quantitative asset manager offering a suite of separately managed accounts and mutual funds. At Newfound, Corey is responsible for portfolio management, investment research, strategy development, and communication of the firm's views to clients.

Prior to offering asset management services, Newfound licensed research from the quantitative investment models developed by Corey. At peak, this research helped steer the tactical allocation decisions for upwards of $10bn.

Corey is a frequent speaker on industry panels and contributes to ETF.com, ETF Trends, and Forbes.com’s Great Speculations blog. He was named a 2014 ETF All Star by ETF.com.

Corey holds a Master of Science in Computational Finance from Carnegie Mellon University and a Bachelor of Science in Computer Science, cum laude, from Cornell University.

You can connect with Corey on LinkedIn or Twitter.

Or schedule a time to connect.