Adventures in TAA

I wrote my first tactical asset allocation system over a year ago (see the alpha version here). Not because I was enthralled with the power of the approach, but because I was concerned that my discretionary trading results would suffer a serious decline following my demise. For the sake of the family finances I wanted something that could function well without me.

Though it shouldn’t be, it’s still a work in progress. It shouldn’t be because even something as simple as the Permanent Portfolio has worked pretty well over several decades, and may perform relatively better as time defeats each wave of “smarter” approaches.

As logical as that is, I can’t do it. I feel stupid using something that simple and have to “improve’ it. So the challenge is to create something that will make me feel smart enough to be comfortable and stick with it, while preventing the additional complexity from having any real impact on the results. Ideally the embellishments should be like the Emperor’s new clothes: Good enough to fool the user, but in reality non-existent.

I had to wait a year for a little turmoil in the bond market, but after some modification I think I have a good candidate. It still needs another year in the sandbox and some tedious QA on the code, but here’s what I’ve learned so far:

  1. For a set it and forget it long-term system, momentum has been, and is likely to remain, the best method for selecting positions. Human psychology has been remarkably stable over centuries, and most people are sheep, so trends are likely to exist as long as humans remain in their current form. Mean reversion will always be around too – but it may stubbornly refuse to revert in your financial lifetime (which leverage in any form will considerably shorten).
  2. The number of simultaneous positions you should hold ends up a function of risk tolerance. Holding one position at a time in a fund switching system generally yields the best absolute returns, but the volatility which creates those returns cuts both ways. It looks great on paper when you know how it all turns out, but when you’re down 35% and staring into the great unknown it feels entirely different. For me, limiting the maximum number of positions to 2 or 3 funds seems the best balance between risk and reward.
  3. However it’s measured, momentum is a trend following approach and is therefore prone to whipsaws, especially in sideways markets. Including 20 highly correlated stock ETFs in your system is going to multiply whipsaws by at least 20-fold as it jumps from one to another, while adding far less than 20-fold improvement in returns.¬† Pick one fund in each asset class (preferably the one with the lowest cost and highest daily volume) and leave it at that. If you can’t hold the line there, then split your money and trade separate portfolios to limit the number of trades between correlated assets. One pot for international, one for domestic, or whatever floats your boat. Putting it all into one bucket will chop you into ribbons.
  4. Cash is an asset class which should be in any TAA system. It’s been almost 40 years since cash was truly king for any length of time, so it’s easy to dismiss the value, but that also means the next coronation is that much closer.
  5. Forget benchmarking. Beating some index means nothing to you as an individual. Your personal goals alone should determine success or failure. Technically, if you are beating inflation by even a little over time, you’re ahead of the game – and over sufficiently long periods I’m not sure much more is possible. If that’s not enough for you, reevaluate your spending and goals. Money is only a means to an end. Figure out what ends you’re really seeking with your money chase. Here is a good one (via Abnormal Returns).
  6. Test your system over as much data as you can and with the worst cases you can devise. Don’t assume the 20th century was a representative sample. In the last part of the 19th century deflation was frequent, incomes and GDP rose at the fastest rate in US history, and the stock market went pretty much nowhere for long periods. If the world has to make sense to you for your system to work well, it’s doomed.
  7. Simple = robust. It’s the fifth law of thermodynamics or something.
  8. Devising a system that’s truly agnostic about the future is extremely difficult. Biases resulting from early experiences can shape a lifetime of investing. So many assumptions are buried in how you view the world and markets, it’s hard to discover and unravel them all. Read every market history you can find to gain a better understanding of the possibilities.

Though successful investing can be described very simply, implementation soon reveals an abyss of chaos and confusion lurking behind those simple ideas. In moments of despair, remember that living below your means will solve the majority of your financial problems given time, regardless what happens in the markets. Your most important task as an investor is to avoid making that too difficult.


The Trouble With Indicators

Market indicators can be like the mythical sirens luring men to their doom. With an irresistible song promising to pick a turning point or signal a trend, they all have the allure of nailing it at some point. But what’s the single best market indicator? Price, plain and simple. It may not be the most satisfying answer, but it’s the only market data that directly changes your account balance. No matter what arcane financial alchemy you can devise, price will always determine its success or failure.

Unfortunately for dreams of easy riches, price is also one of the worst indicators for gaining an edge over the competition. It’s the most widely watched number there is – even your favorite contrary indicator relative knows when the Dow hits a round number. Further, no matter how much fancy math is applied to the price data it can’t increase the amount of information it contains. As a result, the long run results of any system based on price will be quite similar to those of a simple moving average. You really can’t squeeze blood from a turnip.

In trying to overcome these downsides investors turn to indicators like volume, breadth, sentiment, intermarket analysis, and countless others – even economics (a clear indicator of desperation). Though these all add potentially useful information, it comes at a significant cost: They will all hurt your returns. They aren’t 100% correlated with price, so by definition there will be times when price does one thing and they do another, which ultimately leads to additional losses. To avoid that it’s tempting to add more indicators, each attempting to overcome the shortcomings of another, but that’s a sure path to decision paralysis. They’ll most often be contradictory with no clear signal, and in the rare case they’re in agreement they’ll almost certainly be wrong. [1]

What’s the solution? The ideal approach depends on your personality, but the generic version is to change your goals from beating the market to coexisting with it. Buy and hold is almost impossible to beat for mere mortals so stop adding complexity and risk trying to do it. As long as humanity avoids a downward spiral to extinction, a basic moving average system can perform more than adequately to fund a comfortable retirement over a career (see here and here). In many cases outperforming the broad averages on a risk-adjusted basis, while requiring very little time.

That seems like an easy sale, so why don’t more people keep it simple and avoid the troubles with indicators? First, because humans are never satisfied with anything for very long, the urge to tinker and “improve” can become nearly irresistible. Second, individual psychology can make simple trend following approaches difficult (typically due to a need to be right or feel superior), and it can take years of work to alter those preferences. Lastly, “outperforming on a risk-adjusted basis” is really just another way to say “underperform with smaller losses.” Even though that may be the optimal strategy for most people, it can be hard to stomach when trapped in conversation with the latest crop of bull market geniuses who’ve all doubled their money in the last 3 months.

The real money is made by people who are still playing the game 20 or 30 years later. If it takes very low drawdowns to keep you from bailing out, that’s the best approach for you to achieve your goals, even if it means lagging some benchmark to the point of embarrassment. You don’t have to worry about losing assets under management or beating the competition, the happy ending is all that matters. Leave the cornucopia of indicators to the flash in the pan home run hitters. Sustainability is the real key to financial success.


[1] Not because that’s just your sorry luck, but because other market participants will make trades based on far less information while you’re waiting for a consensus, effectively front-running any trades you finally get around to making.

Automation =/= Salvation

Today software is often seen as the solution to nearly all of mankind’s problems, and to an old fart like me, when I’m not swearing at it too much to notice, information technology can indeed seem nearly miraculous at times. But before allowing yourself to be overcome by the siren call of technology as the solution to your problems, here’s something to consider: If you can’t get a slow, manual, human-based process to work well, all that automation will do is allow you to screw up at a much faster pace, with every mistake instantly propagated throughout the system.

The latest software and technology doesn’t guarantee success, and the lack of it doesn’t guarantee failure. The process itself is far more important than the platform and software than runs it, whether it’s a co-located HFT black box or the wetware between your ears.