“Matt’s Breadth Indicator” Update

Happy new year! It’s that time again, when everyone with a blog does a wrap up of the previous year. Here’s my look-back.

Many of you follow along with the “+/-30% per quarter wider-market breadth indicator”. Which is too much of a mouthful, so I’ve humbly named it after myself instead. I wanted to provide an update since I’ve been tracking it for awhile.

The premise is this: breadth from the wider universe of stocks (aka the Russell 3000) can give some indication about the market health as a whole, and can be used to time the purchase/sale of the SPY ETF.

The actual details on how this works were discussed exhaustively with this series of blog posts. If you’d like to feel exhausted too, please give them a read.

First off, here’s a comparison graph of buying and selling SPY using the indicator (no commissions deducted) vs buying and holding SPY, over the period 1/1/2000 through 12/31/2016:

As you can see, the indicator avoided the serious crash of 2008. It also made the bear market of 2000 less sucky. It also kept us out of the market during 2015, which was tumultuous. Equity doubled and drawdowns were much less.

The maximum drawdown of both systems, compared:

max 30qtr DD max B&H DD
-19.71% -56.24%

How did 2016 look? The indicator didn’t suck, but we missed out on some gains. Does that bother me? No, not so much. It’s not going to out-perform buy and hold every single year. Here’s a graph of just 2016:

Below is a graph of the drawdowns for 2016. Obviously, buy-and-hold took a serious hit at the beginning of 2016. The maximum drawdown was -9.15% vs -3.37% for the breadth-indicator system.

This lower volatility doesn’t however make up for the paltry returns compared to buy and hold. It would have however been a lot less scary using the breadth indicator!

Since the maximum drawdown using the system is 2.8 times less than buy and hold over the long term (and very close to that, 2.7x, during 2016), you could make the argument that leveraging your account 2.8 times would give you superior returns with the same risk as buy and hold. This wouldn’t have made up the difference between the two systems in 2016, but over the long term it’s something to consider.

Here’s another, somewhat weird graph to get your head around, so let me explain it. I’ve taken the data from 2000, and (starting at 2001), looked back over a year. I’ve calculated the maximum drawdown that happened during that one year period, and compared the breadth system to buy and hold. I then showing a rolling maxDD for the year prior.

What does this do? It shows you the drawdown you’d experience if you started investing at various points over the last 17 years. Some years, buy and hold has less of a 1-year drawdown than the breadth system.

The parts where the blue line dips below the red line are examples. Doesn’t happen very often.

Some years – indeed, much of the time – B&H shows a rolling 1-year gain/loss % greater than the breadth system. The breadth system however shines whenever a bear market shows up:

When the red line is above the blue line, it means buy-and-hold outperformed the breadth system over the previous year.

I will continue tracking this indicator for a long time to come, because a) it’s a long-term indicator and b) I think it has value, even though 2016 didn’t outperform.

2 thoughts on ““Matt’s Breadth Indicator” Update”

  1. Best wishes for 2017. I enjoy your blog’s content and your clear writing style.

    However I have a question about Matt’s breadth indicator: does the SPY B&H in the charts above include total returns from price and dividends or just price? Unless I miscalculated, the indicator has the investor in cash around 60% of the time. Following the indicator therefore deprives the investor of roughly 60% of the dividend yield.

    I wondered about the effect of missing those dividends. Again, unless I miscalculated: if dividends were taken into account, there would not have been much difference between SPY B&H and 30qtrDD total returns over the period 1/1/2000 through 12/31/2016 (105% and 124% respectively — perhaps a negligible difference after taking frictions into account from taxes, commissions, and bid-ask spreads for trading with the indicator).

    Please let me know if this makes sense. Thank you.

    1. Thanks for your comment, Peter. No, dividends aren’t taken into account here. And yes you’re right, if you’re out of the market with SPY at the wrong time (they’re paid out quarterly), then you miss the dividend. If I compared to a total-return index that reinvested dividends, the results would no doubt be relatively poorer.

      However this isn’t an indicator for investing in SPY per se. It’s an indicator to determine the market health based on a wider index and the breadth measured by it. It’s a conservative indicator in that it generally tells you to get out before a big fall in prices, but stays out of the market a little more than most people would like. I use it not as a method for trading SPY, but as a way to ‘shade’ my other investment decisions. For longer term trades, I move some positions to cash when the indicator looks dicey.

      The indicator performs much better on a risk/reward basis than do traditional indicators like moving averages and the other breadth indicators I looked at. How you use it is of course up to you. 🙂

Leave a Reply

Your email address will not be published. Required fields are marked *