This commentary is available as a PDF here.

Summary

  • We’ve heard from several market participants that they feel that markets have “gotten faster” lately.
  • We define two quantitative measures of speed and examine how speed has evolved since the 1920s
  • We demonstrate that market speed is actually well within normal ranges – but perhaps our perception of normal has changed with the extremely high risk-adjusted return earned in U.S. markets since 2009.


Are markets getting faster?

That’s a common question we receive.  There appears to be a general market feeling that market events are unfolding faster than ever – that to keep up, investors must act faster and trade faster.

Some investors we’ve spoken to posit that it is high frequency trading speeding things up.  Others argue it is the increase of systematic investment methodologies, driving price swings to unnatural extremes.  Others argue that a structural shift in investor risk-aversion post 2008 causes elevated levels of panic.

None of these arguments given, however, are easily quantitatively proven.  They are gutfeelings that come with little-to-no evidence.

To actually test if markets are getting faster, we need to come up with a measurement of speed.  What we believe most people mean when they say markets are getting faster is that prices are covering a greater distance in a shorter amount of time.

To systematize this idea, we identified local price highs and lows using a 10-day simple moving average.  Local extreme points are identified every time price crosses above or below the moving average.  The distance covered between these extreme points, and the time it takes to cover that distance, is how we will measure speed.

Measuring Market Speed

Using market data from the Kenneth French library (NB: this is a total market measure, not just large caps), we can then find the average daily speed per quarter going back to 1926 and plot it to determine if markets have been moving at a faster pace.  The answer?

Definitely not.

Avg Speed of Price Swing per QuarterSource: Kenneth French Data Library.  Analysis: Newfound Research.

Perhaps averages aren’t the right measure.  Maybe we have to look at the extremes and ask, “are they getting more extreme?”  So we can plot the maximum speed, per quarter, as well.

Max Speed of Price Swing per QuarterSource: Kenneth French Data Library.  Analysis: Newfound Research.

Certainly higher, but not unusually high.  2011 was actually worse.

So perhaps we have the wrong measure.  Perhaps it is not the speed of the drawdowns, but rather the number of large drawdowns themselves.

So let’s plot the number of days in between 10% drawdowns.  If drawdowns are happening faster, we should see fewer days between them.

Number of Days since Last 10 Pct DrawdownSource: Kenneth French Data Library.  Analysis: Newfound Research.

What do we see?  As it turns out, we just went through an incredibly prolonged period of no 10% drawdowns.  August 2015 broke an 800+ trading day streak.

If anything, we just proved the opposite.

What if we take a broader view of drawdowns, though?  What if it isn’t just about the magnitude, but also the frequency?  We can use a measure called the Ulcer Index to quantify that! (In this case, we’ll use a 252-day lookback).

Yet once again, we can see that we’re much closer to all-time lows that all-time highs.


Ulcer IndexSource: Kenneth French Data Library.  Analysis: Newfound Research.

Another way to look at this data is to look at the number of significant price swings we’ve had over some trailing period.  Below we plot the number of 10% swings (whether positive or negative) over rolling 3-year periods.  We see a significant drop-off after 2011 (once 2008 falls out of the picture) and a continued decline to historically low levels.

Number of 10-Pct Moves Rolling 3 YearsSource: Kenneth French Data Library.  Analysis: Newfound Research

.
The four 10% up/down moves over the last 3 years is less than half the long-term average.

Now, for momentum-based models such as our own, not all 10% moves are made equal.  Seeing a number of 10% moves up in a row is much better than seeing reversing moves up and down, causing a choppy sideways market.  So one way we can evaluate the environment the consistency of these moves is by looking at a running total.  We add +1 when there is a positive 10% move and -1 when there is a -10% move.

Time-series momentum users will want this line to either be going up or down consistently.

Cumulative Tally 10-Pct MovesSource: Kenneth French Data Library.  Analysis: Newfound Reesarch.

We believe the cumulative trend is still exhibiting very normal behavior.

So what’s going on here?  We would posit that investors are conditioned by their environment – and the recent environment has been easy.  How easy?

Rolling-Sharpe

Source: Kenneth French Data Library.  Analysis: Newfound Reesarch.

From 3/16/2009 to 12/31/2015, the market exhibited a realized Sharpe ratio of 1.06.  By comparison, the long-run Sharpe ratio of the market from 1926 through 2015 is 0.36.

Compared to other 1712 trading-day periods, that puts the recent bull market in the top quintile – outdone only by the dot-com boom and the post-World War II boom.

Markets aren’t getting faster; we’ve probably just been lulled into a false sense of security lately.

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.