Wednesday, February 17, 2016

Combining Momentum and Dollar Cost Averaging for Smoother Results

Josh Brown (i.e. The Reformed Broker) recently shared the aptly titled post How to Make Volatility Your Bitch highlighting how dollar cost averaging into a volatile market can lead to higher overall returns:
Door number one – you spend 15 years putting $1000 into an investment every month for 15 years, with the possibility of seeing that investment get cut in half twice.
Door number two – you spend 15 years putting $1000 into an investment every month for 15 years, with the same annual performance of what’s behind door number one, but no drawdowns.
Which would you choose? 
On the surface, you’d choose door number two. Of course, who wouldn’t? 
But it’s the wrong choice. The trick here is to remember that you’re adding to the investment at a rate of $1000 per month. That’s when you realize that door number one, with it’s twin 50% crashes, is the better option.
His point is an important one for long-term investors... you would rather pay less (than more) for a security today if it is worth more in the future and for long investment horizons that has typically been true. So in general, regularly contributing to your retirement (or other long-term goals) is good practice.


The Caveat

BUT there is a caveat... dollar weighted returns look great if / when the dollar weighted price you paid was lower than the price at the ending date. As dailyVest outlines (bold mine):
In contrast with a time-weighted approach, the dollar-weighted rate of return calculation method does measure the size and timing of cash flows, as well as the investment performance. Thus,
  1. Periods in which more monies are invested contribute more heavily to the overall return – hence the term “dollar-weighted” 
  2. In this case, investors are rewarded more for larger investments made during periods of greater price appreciation 
Given we are so close to the all-time high in the S&P 500, chances are each dollar invested over the last 15 years was below (or well below) the current price, resulting in more investments having greater price appreciation and dollar weighted returns > time weighted returns. So... rather than looking at just the current 15 year period, let's go back and look at 15 year periods ending 5 and 10 years back (which end at less of a peak) to see how well the same door number one vs door number two worked out.

Note: the analysis below shows $1000 invested each month in the S&P 500 and in a return stream with identical 15 year time weighted returns, but with 0% volatility.


Dollar weighted returns > Time Weighted Returns over Most Recent 15 Years



Dollar weighted returns < Time Weighted Returns over 15 Years Ending 2010



Dollar weighted returns < Time Weighted Returns over 15 Years Ending 2005

In these examples we see just how important the ending point is in determining which return stream "wins", as well as how important the end date is in the overall growth of the $180,000 contributed (which is a reason why investors generally should derisk as they approach retirement). It also outlines why dollar cost averaging into a solution that can protect against the downside may be beneficial relative to a buy and hold strategy by limiting the amount of dollar weighted contributions made at poor entry points.


Dollar Cost Averaging in a Capital Preservation Strategy: The Case for Momentum

Because I have the urge to compare all buy and hold strategies with momentum, the below replicates the above charts and adds a momentum equity curve with this simple rule at month-end:
  • If S&P 500 > 10-Month Moving Average, then S&P 500
  • Otherwise, Aggregate Bonds
Similar to what is shown in the charts above, the charts below outline the growth of $1000 a month into the S&P 500, an identical time weighted return series with zero volatility, and the above momentum strategy.




The momentum strategy provided much more consistent dollar growth in all three time frames and in these specific windows materially outperformed both a buy and hold and 0% volatility iteration (this will not necessarily be the case in all periods - especially in up and to the right equity markets). So... perhaps it's the combination of consistent contributions and a strategy more focused on capital preservation that can more easily make volatility your bitch (without the volatility).

Wednesday, February 10, 2016

Avoiding Bear Markets to Improve Risk-Adjusted Returns

Ben Carlson of A Wealth of Common Sense has a recent post, When Global Stocks Go On Sale, outlining that it is typically a pretty good time to be buying when the MSCI World stock index is in a 20% or greater drawdown.

His insightful takeaway and chart outlining the historical drawdowns and forward performance of the index is below:
There were only two times out of the ten bear markets where stocks weren’t higher one year later. Only once were stocks down three years later. And there was never a period where stocks weren’t higher five years after initially falling 20%. The paradox of investing is that the best times to put your money to work are often when things seem like they’re never going to get better.

While I in no way disagree with his insight, especially for a buy and hold investor thinking of selling, I thought it would be fun to share the completely opposing strategy that avoids these periods of distress, as well as one that avoids stock exposure after even one month of negative performance.


Avoiding Extended Drawdowns May Improve Risk Adjusted Returns

As I outlined in a previous post, Using "Normal" Drawdowns as a Timing Signal, an investor who sold their S&P 500 allocation whenever the S&P 500 index was in a drawdown of 10% or more, and instead held bonds, had similar long-term returns as a buy and hold investor, but with materially less risk and drawdowns.

A similar situation has played out for investors allocated to stocks within the MSCI World index when drawdowns were less than 10% or U.S. treasuries when the MSCI World was in a drawdown greater than 10%, while a 20% threshold wouldn't have held up quite as well as the 10%, but would have provided roughly similar returns with less risk than a buy and hold investment.



What gives?

The reason for the improved risk-adjusted performance has been the power of momentum within the MSCI World index during drawdowns. When the MSCI World index ended a month at a roughly 10% drawdown, it often moved lower... sometimes much lower. At a 20% drawdown, only 2 of the 5 times was this in itself a decent short-term buying opportunity (highlighted in green). The other 3 times presented a better opportunity further down the road.



Buying Only at the Peak

Taking this "drawdown avoidance" to the extreme, let's see how an investor in the MSCI World index would have performed if they only bought when it was making new end-of-month highs. In this example, an investor is only holding the MSCI World index if the previous month was at an all-time high, otherwise U.S. Treasuries.


While there were long periods of relative underperformance (and this is an extremely high turnover strategy), the resulting performance and lower risk offers some insight into how a strategy that is less exposed to risk, yet can avoid loss of capital, may actually be able to improve absolute and relative performance.