What level of returns is impressive? That question begs the questions, “impressive to whom?” and “over what timeframe?”

I would have to say I’m answering the “to whom” question from the standpoint of intermediate-to-longer-term holders, traders who hold positions for weeks/months all the way up to year+ periods, and who aren’t *incredibly* active (i.e. not turning over the whole portfolio every month); this is the target for most newsletter-writers and the usual position length and turnover amount for many fund managers that aren’t into stab art, er, stat arb, trading. It ain’t quite “lazy, buy-and-rebalance” trading, but it ain’t “watchin’ the markets every minute” trading, either; at one end of the spectrum are the do-it-yourselfers following a service or spare-timing it after hours, and at the other end of the spectrum are the fund managers running money in a fairly conventional way. I exclude day-traders and options traders for several reasons, including the magnifying impact of high activity and the difficulty in getting quality aggregate statistics for their returns (more on that later).

Over what timeframe should the results be evaluated? There’s no easy answer, so where possible, I’m going to go for the round numbers of 3-, 5-, and 10-year annualized results, in my search for “impressiveness.” Anybody with a little skill can get a 60% to 80% number in a booming bull market recovery like 2003-2004, but can they swing long-term numbers? In the long run, long-term numbers are what counts. The problem with picking a timeframe of returns is that the same trader or fund manager can have different results over a long time period, leaving us to wonder which is the real result. If you were to check out a real-life example recent eight-year return period, ask yourself: Which 3-year result is accurate, 2002-2004, or 2005-2007? Is the 5-year number from 2000-2004 more “accurate” than the 5-year number from 2003-2007? This is the inherent problem with selecting a timeframe for evaluation.

Another inherent issue with timeframe, at least, when evaluating OPMMs (other people’s money managers), is that their objectives/techniques/systems may have changed over time, especially if their AUM (assets under management) have grown. It’s hard to scale a strategy from deploying $6.5 million to deploying $650 million and still hit the same CAGR (compounded annualized growth rate) targets. This is something to keep in mind, even though I’m not going to explicitely control for it in this post.

With “to whom” and “over what timeframe” answered, at least for the purposes of this post, I started looking for publicly-stated return streams to evaluate.

Hedge funds have some significant biases in their reported returns, as noted by the atrociously-named (in my opinion), but otherwise very worthwhile site, All About Alpha. There are also some barriers to retail scrubs (i.e. non-accredited investors) like myself being able to track down accurate aggregate returns for the purpose of evaluation. Hedge funds are also notorious for “home run hunting” and playing the game based on their high-water marks; in other words, they have some powerful incentives to swing for the fences, and simply close down when they go bust. This leads to misleading returns, to say the least, where a fund may return in the 40% range for several years before the “Widowmaker Spread” shuts them down, leading their traders to go to other funds (and the fund managers to open new funds) where all the investors are subject to high-water marks that are still achievable. Oh, bother. I’m excluding hedge fund returns from evaluation for this reason.

Some of the other asset managers, such as Clear Asset Management, ostensibly wanted to focus on individuals, but often wind up marketing to institutions. If you want their performance numbers, which Roger characterizes as “top-tier,” “winning performance awards … again and again,” and “top-performing,” you’ll have to call them, or fill out a form, and hope they respond. Or, you could get an anonymous email from a friend who helped out! Hint, the three-years-and-a-couple-of-months compounded annualized number for their oldest long-only funds are all right around 20% annualized, with a good spread (from 0%-ish to 30%-ish) on the funds’ returns for the 2007 calendar year. It’s always nice to know, when a long-term industry vet says a return is “top-tier,” exactly what that return is, which I why I wanted to highlight that fund (and that post) as an anecdotal example.

Newsletters don’t generally self-report their returns, but there is a service that does the calculations for you: the Hulbert Financial Digest, which has some newsletters they’ve been tracking for a quarter of a century. Over the last 25 years, a time period where, with some burps here and there, this generation has seen the market go up, the top-performing newsletter has been “The Prudent Speculator,” with 16.7% returns annualized. Only two of the nineteen tracked over that time period beat the DJ Wilshire 5000 index. Over the last ten years, the top-performing newsletters have ranged from 13.6% to 15.7% annualized (out of 64 tracked for that period). Over the last five years, it’s no surprise that the top-performing newsletters had a foreign-market focus, with returns ranging from 28.3% to 37.2% annualized (out of 132 tracked for that time period). I don’t have three-year numbers handy, so I’ll have to leave it at that. It seems, from a glance at the newsletter performance, that longer-term numbers in the 20% range might, indeed, be impressive.

Mutual Fund returns are easier to get to, although the various screeners all seem to have different categorizations to put them in. I went to my screener of choice, and did a default search with everything left blank, just to see what the size of the field was (for this screener). I got 10,179 funds returned, a sizable amount. Based on the 20%-ish range established above, I decided to look for the funds that had 20% or higher annualized returns over the last three, five, and ten year periods. I left all the other fields (load, expense, open or closed, AUM, risk rating, etc.) blank. I got only 13 funds returned. Hmm, 13 out of 10,179 is about 1/8th of one percent. Maybe 20% is impressive over the long haul! Broadening it out to 15% annualized over those time periods, I saw only 66 funds out of 10,179 - or about 2/3rds of one percent.

The great thing about Commodity Trading Advisors (CTAs) is that they have a website which publicizes their funds returns; the Institutional Advisory Services Group (IASG). It’s easy to find funds sorted by three and five-year returns at their site, but I haven’t gotten it to regurgitate ten-year numbers for me, yet. Another important thing to remember about CTAs is that they’re playing with futures contracts, and there’s an inherent amount of leverage available. For example, the NYMEX just last week raised crude oil margins to $7,763 for customers, meaning that, at $100 a barrel, you, the joe blow retail punk, can swing 12.88 to one leverage in an overnite position. Members have 15.8 to one leverage available, BTW. This is without borrowing a penny! Some of the futures contracts have higher implied leverage, up to the low 20’s. This high amount of leverage tends to make the range of returns more dispersed, with higher (and lower) returns appearing more frequently than they do in unleveraged stock instruments like mutual funds.

In my search for CTAs with three-year annualized returns over 20%, I found 74 out of 330 ranked. When looking for five-year annualized returns above 20%, I found only 34 out of 330 ranked, pointing out (again) the issue of “what timeframe” as being key to evaluating returns. I brought the data to a spreadsheet to look for those managers with over 20% annualized in both three- and five-year periods, and found only 27 of the 330. So even with the extra leverage of playing with futures contracts, fewer than 10% of the professional money managers could crank out multi-year consistent returns above 20% annualized.

Other than the newsletters, which are aimed at the retail crowd, most of the return streams I’ve looked at above are “professional” or institutional returns. For the retail crowd, having access to smaller capitalization stocks and less capital to deploy can be an advantage, but the disadvantages are higher transaction costs and less access to leverage and information, as well as some limited opportunities on other fronts. I’d guess it evens out the playing field, provided each one sticks with their relative advantages.

It seems that grabbing numbers in the high teens or above, over long time periods, is a relatively impressive achievement.