Dumb Alpha

Today’s post looks at a series of old articles by Joachim Klement.

Dumb Alpha

Joachim Klement

We’ve come across Joachim Klement many times before.

  • He works for Liberum and writes his own blog/Substack newsletter.

Back in 2015, he began a series of twelve articles for the CFA Institute on the subject of dumb alpha.

  • The title was borrowed from Brett Arends of MarketWatch.

To quote from the first article:

Modern finance constantly busies itself with the development of new, more sophisticated ways to manage risk and generate returns. On the opposite end of the spectrum are simple ways to invest that have a proven track record of providing superior investment outcomes.

The point of the series was to look at investment techniques “so simple it is surprising how well they work”.

  • In today’s post, we’ll start working through the highlights from that series.
Asset Allocation

The first article looks at asset allocation.

Given the uncertain nature of future returns, Joachim identifies two logical responses:

  1. Everything in safe (low return) assets
    • Unfortunately, this won’t protect you from inflation, which is more relevant than when the article was written
  2. Invest the same amount of money in each asset class (however, you define them)
    • The simplest option would be stocks/bonds, but a fairly split might be stocks, bonds, cash, real estate and commodities/gold
    • Or you apply the same principle to stocks and instead of a market-cap-weighted index fund, invest an equal amount in each of the 500 stocks in the S&P

Approach number two is surprisingly successful:

In their 2009 article “Optimal versus Naive Diversification: How Inefficient Is the 1/N Portfolio Strategy?,” Victor DeMiguel, Lorenzo Garappi, and Raman Uppal tested this naive asset allocation technique in 14 different cases across seven different asset classes and found that it consistently outperformed the traditional mean-variance optimization technique.

Over a full business cycle, this approach outperforms in terms of returns, risk-adjusted returns, and drawdowns.

  • Unfortunately, since the 2008 financial crisis, stocks (and particularly large-cap tech stocks) have massively outperformed other asset classes, so nobody wants to hear about diversification.

Joachim suggests at least using an equal-weighted portfolio as a benchmark.


Equal weighting is the starting point for the hand-crafting process (so named by Rob Carver) that I use for my passive portfolio.

  • Deviations from equal weights need to be justified by higher expected returns, lower volatilities or inter-asset correlations).

And I don’t try to fully optimise to a short back-test, preferring a loose optimisation to date from more than 100 years.


It is pretty clear why this dumb alpha works. Putting the same amount of money in every stock systematically prefers value and small-cap stocks over growth and large-cap stocks.

So this is a form of factor investing, at least within stocks.

  • The other advantage is its resilience to forecasting errors, compared with more sophisticated approaches.

And because everyone is terrible at forecasting, making no forecasts whatsoever proves to be a significant advantage.

Forecasting

The second article is about forecasting.

  • Joachim thinks that the finance industry’s approach is inferior to that of physics, where he used to work.

While our industry is obsessed with forecasting, it is astonishing how little attention is paid to the accuracy of forecasts and their estimation errors.

He looked at research by Markus Spiwoks et al which compared professional analysts’ forecasts to a naive prediction – based on the assumption of markets following a random walk – that the best predictor of a value one year into the future is the value today.

See also:  Do Short Trackers Work As Hedges?

The first (and only) positive finding is:

Only a few analysts [around 10%] seem able to consistently outperform the consensus forecast compiled from many  different analysts.

This is the “Wisdom of Crowds” effect.

  • Unfortunately, none of the expert forecasts beat the naive forecast.

Your best bet is to assume that the future will be like the present.

Mean reversion & compounding

The third article argues that long-term forecasts are also poor because mean reversion cannot overcome the effects of compound interest.

While return forecasts can be widely off the mark in any given year, in the long run, returns should converge towards a rather stable long-term mean. Because of mean reversion, it should be easier to forecast long-term returns than short-term returns.

Emphasis on the word should.

  • In fact, Pastor & Stambaugh found that estimation errors in early years propagated too much for mean reversion to overcome them.

Joachim quotes an investment with a 10-year average annual return of 10%:

If in the first year the return is -10%, the average return over the subsequent nine years needs to be about 12.48% per year to make up for this shortfall. A 20% estimation error in the first year requires a 24.8% increase in annual returns over the next nine years.

Similarly, if the first year is 0% (10% error), the other nine years need to be 11.17% (11.7% increase).

The investment results of the first few years have an oversized influence on the long-term investment returns.

This is called sequence risk and is more commonly discussed in the context of decumulation (retirement).

  • Sequence risk doesn’t apply to steady-state portfolios (with no inflows or outflows), but few of these exist in the real world.

In retirement, investors often protect against bad losses in the early years by lowering their equity exposure (this is known as a bond tent).

  • The reason that sequencing risk is less discussed in accumulation is that contributions tend to increase with age, and inflation increases the nominal size of the retirement pot.

These two effects mask the impact of sequencing risk, but it remains the case that the returns of the first few years have a disproportionate effect on the size of the terminal pot.


Joachim suggests a couple of ways to mitigate the fact that the uncertainty about future equity returns does not decrease in the long run:

  1. Equal weight, minimum variance and risk parity portfolios do not rely on forecasts and should outperform traditional portfolios.
  2. Resampled efficient frontier methodologies or Bayesian estimators can include estimation errors in the portfolio construction process, building resilience to unexpected events.

Private investors should probably stick to the first option.

Momentum

The fourth article looks at momentum, the tendency for prices to go up simply because they have gone up in the past.

  • Joachim admits that he found this a dumb idea for a long time, preferring value and contrarian investing, which both bet against the crowd.

However, momentum works, at least in the medium term:

While winning investments of the last three to five years tend to underperform as mean reversion kicks in and winning investments of the last month tend to  underperform as well, winning investments of the last three to 12 months tend to outperform in the subsequent months.

Although many dozens of systematic anomalies in asset returns have been found, many appear to be the product of data mining, with value and momentum the most robust.

See also:  Renaissance - Too Good to be True?

The problem with momentum investing – and trend following – is that it is self-perpetuating (pro-cyclical) and tends to result in a bubble followed by a crash/drawdown.

  • For this reason, it is best used as an overlay on a traditional (passive, long-only) portfolio so that the opposite convexities of the two approaches combine to lower volatility.

Joachim recommends taking advantage of momentum acceleration, as described by Didier Sornette et al.

They calculate a simple measure of past change in momentum — for example, the return over the last six months minus the return over the preceding six months. Stocks with the highest acceleration tend to have higher returns. 

And the returns from an acceleration strategy are higher than those from a plain momentum strategy.

  •  Joachim interprets this as getting in on new trends early and perhaps getting out as the trend decelerates.

That’s it for today – we’ve covered four of the twelve articles, so I anticipate that there will be another two posts in this series.

  • Until next time.

Mike is the owner of 7 Circles, and a private investor living in London. He has been managing his own money for 39 years, with some success.

You may also like...

Leave a Reply

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

Dumb Alpha

by Mike Rawson time to read: 4 min