January 10, 2013
Context Is (Still) King For Portfolio Design & Management
It happened again. I was reading a story about the apparent hazards that are expected to derail an asset class that's in my portfolio and I thought, gee, I better sell. The article presented a strong case for seeing red in the near future. But then I remembered that the asset class is just one piece of my diversified portfolio, and that my rebalancing strategy will take care of any extremes in the various components. Thinking about the big picture for my personal asset allocation reminded me that the article wasn't all that applicable to my situation after all. Once again, my initial emotional reaction turned out to be not so helpful after all in money matters.
Stories about the alleged opportunities and risks for a given asset class at a specific point in time are a staple in the financial media. Some of them are actually right, but not always. I should know, since I've been known to write a few of these gems through the years. But every time you read one of these articles you should remember that the information is almost always presented in a strategic vacuum in terms of your portfolio. That's inevitable, of course, since only you know what's in your portfolio, and so only you know how much relevance, if any, applies when it comes to current news on a relatively narrow slice of the world's capital and commodity markets. In other words, analyzing assets in isolation of your total portfolio can lead to bad investment choices.
The caveat wouldn't mean much if we had something approaching near certainty that the analysis du jour for a given asset class was accurate. In that case, we could throw all the standard rules about risk management out the window. In the real world, of course, imperfection infects every effort to peel away the cloud of unknowing on the morrow. Fortunately, there's a reasonably effective solution: asset allocation and rebalancing.
When you hold a broadly defined portfolio of the major asset classes, you're always prepared to exploit return volatility, regardless of the source or whether it's a total surprise or widely anticipated. Once you wrap your head around this idea, many of the updates from the usual suspects lose their relevance for your portfolio.
We can and should debate the details for structuring a portfolio in terms of defining asset classes, how many to own, etc. But the key point is that in order to harvest risk premia from price volatility, you first must own a broad set of assets. That means owning a mix of winners and losers on a regular basis, and not getting too worked up over any one piece of the portfolio at one point in time. Ideally, you'll own an expansive set of assets that maximize the supply of low and negative correlations across the portfolio--an essential feature for profiting from rebalancing opportunities.
Yes, it all boils down to buy low and sell high. Duh! But obvious lessons all too often remain elusive for investors on the road to earning a decent return. One reason is there's a tendency of going off the deep end in deciding that a certain asset class should be avoided, or that another asset is a sure thing and so it's time to overweight in the extreme. The problem is that if you spend any time studying the historical record of all the asset classes, in context with one another, it's clear that surprises are a constant. The implication: own everything and focus like a laser beam on managing the mix.
To take a recent example: US Treasury bonds. How many times over the last several years have you heard that they're in a bubble? The countless warnings sounded reasonable. But did you also know that Treasuries have delivered handsome returns in recent years? The iShares Barclays 20+ Year Treasury Bond ETF (TLT), for instance, generated a 14% annualized total return over the last three years through January 9.
What should you do now? The same thing you should have been doing all along: diversify and rebalance. If you've owned Treasuries for, say, the last three years, you should have been rebalancing the portfolio periodically. If so, you've been harvesting some of the tidy gains from government bonds recently and in the process kept a lid on the Treasury allocation. Same old, same old. It's not exciting, but it works.
Suffice to say, you can always find smart people telling you to buy, or sell, just about anything. But stellar track records based on this advice, alas, tend to be the exception.
Don't misunderstand: it's essential to sell overpriced assets at some point and favor the underpriced ones. In fact, that's everyone's goal. Results, of course, vary dramatically. One way to systematically boost the odds of earning average-to-above average returns relative to the crowd is to diversify broadly and rebalance periodically. The first phase of this one-two strategy is easy: buy a varied mix of ETFs, for instance. The second phase is tricky, although you can probably do quite well with a simple rebalancing strategy of, say, returning the asset weights back to pre-set levels once a year, or something along those lines. This is one simple way to keep bubble pricing at bay, while also making sure that you'll also ride the wave when the today's out-of-favor asset classes mounts a revival.
Truth be told, there are numerous intriguing possibilities for designing rebalancing rules to juice performance, reduce risk, or both. But even the world's greatest rebalancing strategy faces serious headwinds in a portfolio that holds too few asset classes. Keep that in mind the next time you read an article that seems to hold all the answers about one asset class. Context is critical in portfolio design and management, even if it's routinely overlooked in today's hot investing story.
Posted by jp at January 10, 2013 6:26 AM
Good point, which of course raises the question of what's the definition of "too few" asset classes? It's a gray area, in part because the past is only a fuzzy guide at best to the future. There are certainly historical periods when a relatively small handful of asset classes outperformed a broader set, and vice versa. The same is true with rebalancing more, vs. less, or even not at all. The problem is that it's not clear what's going to work best.
I don't think a slavish allocation to everything is absolutely necessary, but the more you move away from GMI, the bigger the bet. That's fine if you're confident in your bet. If you want to beat the market, you have to hold something different from its opportunity set and/or in different allocations. But once again, positive alpha in the world is offset by negative alpha in active bets, including active asset allocation. Caveat emptor.
In any case, I'm aware of the Permanent Portfolio and I'll take a fresh look at it and perhaps post some comments soon. One focus is to compare it with Vanguard's STAR fund. There are many multi-asset class funds with long track records. Overall, the results are middling. The issue, of course, is picking the above-average performer for the years ahead, if you're paying above-average fees. Alternatively, you could opt for a low-cost mix and boost the odds of earning average-to-above-average results.
Posted by: JP at January 12, 2013 6:54 AM
Jim- Intrigued by your statement "But even the world's greatest rebalancing strategy faces serious headwinds in a portfolio that holds too few asset classes." The HB Permanent Portfolio with only 4 asset classes seems to have weathered just about everything thrown at it, including 2008, and returned 9% CAGR over the last 40 years with excellent Sharpe ratios. This would seem to compare extremely favorably with GMI and its 13 asset classes. Would be grateful for your take and what you see as the comparative advantage of GMI over PP.
Posted by: davidr at January 11, 2013 3:00 PM