When you compare one fund to another, most people just look at their returns. (Total returns over a period of time, or equivalently, CAGR over the same period of time. We won't even mention the phony "Average annual returns" measure that mutual fund companies use when they advertise.)However, that's not the whole picture. Some funds only achieve their high return by taking on excess risk. So you have to also consider some measure of the riskiness of the fund, however you choose to define it. Modern Portfolio Theory looks at derived statistics like volatility, Sharpe Ratio, and a whole buncha Greek letters. They even come up with something known as "Risk-adjusted return" -- which I guess is something similar to "If we held X dollars of this fund and $10,000 - X of cash so that it had the same risk (uh, volatility) as the total market, what would the return be?"This is too complicated for some people, and there are various ways of trying to simplify it. Some websites show a "speedometer" style graph showing whether a fund is less risky, as risky, or more risky than the appropriate total market. Others try to break things up into boxes. Morningstar has the widely known "Star" system where each fund is awarded 1 to 5 stars based on both its performance and risk during some specified past time period.Why is risk something to be avoided? Well, the risks that worked out for them last year are not going to be the same risks that work out for them next year. The market giveth and the market taketh away. Folks who are withdrawing from their portfolio need to be especially risk averse because any drop in their portfolio causes them to drain their reserves faster.Even people who aren't forced to withdraw should use some risk management, because of the emotions involved. Most people can't viscerally handle the sight of 50% of their money just disappearing (even if it was after a 100% gain).Recently we were discussing Joseph Shumpeter's quote, and we concluded that for someone withdrawing from his portfolio, the only safe withdrawal rate is the risk-free interest rate, even if the portfolio has a higher return. Down times will eventually come and he need to have something in reserve for those times. If he spends the whole amount of the increase, he'll be sorry later.Back when I first started investing, my thought was "Get the riskiest legitimate investments I can find, since even if I get wiped out, since I don't have much in there, and I'm contributing a lot each year, it won't put off early retirement by that much time." But I realized that it was hard to tell risky investments from just plain phonies. I steered clear of the NASDAQ (good thing, too...I was way off on my forecast of the dotcom crash).[Funny. I still haven't invested a lot (my net worth is less than my salary) but now I am already seeing the "preserve wealth" instinct kick in. If I'd known in 1999 I'd have had this much in bonds and cash in 2002, I would've been shocked.]Anyway, back then I was contemptuous of the very concept of "risk-adjusted returns". My logic ran, "You can't spend a lack of risk." Well now thanks to the Shumpeter discussion I know that's faulty logic. When you're spending, lack of risk is very important!However, is this necessarily the case for someone who is not spending, but is constantly adding to the portfolio (perhaps with an eye to keeping percentages in balance)? I'm still at the stage where the money I add to the portfolio vastly outweighs any internal increase in the portfolio's value.So, I am once again searching to see if risk-adjusted returns are meaningful to me. I suspect they're not, but I'd like to know why.
you have to define "risk", and each of us defines risk in a unique way. this is what martin whitman means when he says that risk is meaningless without an adjective in front of it.for me,capital risk = permanent impairment of invested capital.permanent = a drop in the value of our capital that is unlikely to be regained within a significant time period at an estimated CAGR.significant time period = 4-5 years.you can plug any CAGR or time horizon you want into that definition and calculate a capital risk best suited for you. in the way i define capital risk, risk adjusted returns are paramount because impairment of capital detracts from your net worth. it is not something you surmount. it is an impairment.for example, lets say i have a client with a cost of capital of 10%, and a time horizon of 5 years. for them a 40 % drop in value of capital invested is a permanent impairment because it can not be "made back" in 5 years at a CAGR of 10%. LTCM and Enron clearly show why risk adjusted returns are important. risk adjusted returns are important because of ergodicity. as nicholas taleb points out in his book, he believes that most successful traders who blow up do not blow up "by accident", rather they did not have a viable approach in the first place. "they were merely deriving their income from the small probability of a large loss (up the escalator down the chute)" over a limited time span until ergodicity did them in.you see the opposite effect in casinos where the equation is flipped upside/down (low probability of large gain). someone will come into the casino, hit it big early, then proceed to lose all of the gain and more as the statistical distribution reasserts itself with time.more later.tr
I think you can approach this issue in 2 basic ways: 1) the "EMH/MPT" approach2) the "EMH/MPT is BS" approachIf you take the EMH/MPT approach you essentially diversify away a lot of separate risks, which can perhaps now be lumped together as "company- or sector-specific risks". You do this by holding > 30 "random independent" companies, and your portfolio is basically going to move in concert with the underlying universe from which those companies were drawn (e.g. S&P 500, Wilshire 5000, etc.). Based on historical returns of equities, bonds, and cash, you can add in an appropriate weighting of bonds and cash to dampen the maximum potential loss that you're psychologically prepared to take in any given month, quarter, year, decade, etc.You have basically accepted mediocrity, less friction. And I agree with Bogle that you should probably focus on minimizing frictional costs, because there's not much else you can do to boost returns. If I was going to take this approach (and I'll have to soon, when I start a 401-k), I'd use Vanguard Funds or other exchange-traded funds to minimize MERs and I'd have minimal turnover, but I'd probably take a contrarian approach (e.g. tactical asset allocation) with new contributions. I wouldn't dollar-cost average into bonds or stocks, I'd purchase money market funds instead and then make tactical purchases of stock index funds or bond funds when prices were low relative to recent trends (i.e. breaking the lower Bollinger band, as based on weekly prices). With the "EMH/MPT is BS" approach, which I utilize nearly exclusively, you rely on sufficient though not excessive diversification. For me, this means surviving a company specific blow-up like Enron or USG with a tolerable level of capital impairment. Being long 10 companies is a typical level of diversification for me, but it can be as many as 20 at times. And I could tolerate as few as 6, and would probably be a better investor if I did, but I get bored too easily.There are very few mutual funds that practice this type of investing, and fewer that do it well, but Third Avenue Value (TAVFX) and Longleaf Partners come to mind in the U.S., or ABC Value in Canada where I'm from.Adopting this approach means that you have to have a certain immunity to "pricing risk", the risk that other investors will continue to mis-price an equity. If I believed TKIOY was undervalued when I bought it at $36, I ought to believe it's even more undervalued when it sinks below $32. When I'm way underwater on a position, I'll begin to think in terms of "there's an increasing probability that others know something I don't know", but I generally don't react to price risk per se (for example, as you'd have to do if you were carrying positions on margin).I don't pay any attention to beta, or implied volatility, other than to recognize it for what it is--the market's recent and future expectation for a given stock. If I think volatility is high, I might try to sell it with a covered option. If I think volatility is too low, I might structure a position with options instead of with shares. But it doesn't ever change my underlying opinion about the desirability of equity ownership in the underlying company.I'm usually always hedged with 4-6 short positions. This isn't because of a conscious effort to be beta-neutral, although I certainly feel better knowing that a 300 point smack-down in the S&P probably isn't going to affect my portfolio by more than 5%. I enter into short positions for the same reason I enter into long positions, because I expect to make money on them. But rather than thinking I'm the great contrarian, I make a conscious effort at times to imagine I'm just another lemming.You know, in trying to explain what I do, I get the uncomfortable feeling that I sound a lot like one of those blokes going up the escalator, but down the chute. But I don't think so. I'm drawn by another quote of Taleb's: "The best hedges are those you are the only one to put on." Man, I'd sure win an Oscar for that one.Todd
I define risk as "time until early retirement". For me the operative number is not the number of dollars in the account, it's the time (or maybe the reciprocal of the time, or (my_life_expectancy - the time).My point being that risk-adjusted returns are meaningless for me because they measure only volatility, which is an entirely different risk than the one I'm interested in.No particular time horizon. That's the dependant variable in my situation, not a parameter.A 40% drop in the value of my portfolio this instant just means I need to work an additional year. It isn't permanent.How does one know the probability of a large loss? LTCM thought they had a surefire system, and they were making good money for a while. They edidn't think they were acting in a risky manner; they thought they had their positions completely hedged and were just profiting from market inefficiencies. They were experts at risk measurement and management....or so they (and the Nobel committee) thought.IMO, Enron was (and is still) primarily a vehicle for "artists" to conceal assets and debts. [Didn't you see the report in the paper today that Enron North America, declared bankrupt, has been transferring billions of dollars out of its accounts, to points unknown?]In neither of those situations was it possible to calculate the risk beforehand. Again I say: Why should I pay any attention at all to risk-adjusted numbers?
"The best hedges are those you are the only one to put on." Didn't Solasis just say a month ago "The market works to make useless the greatest possible volume of hedges"?
When you're spending, lack of risk is very important!However, is this necessarily the case for someone who is not spending, but is constantly adding to the portfolio...?If you plan to periodically invest money in a mutual fund and you have a choice between two funds that both have the same total return as the market over time, but one is more volatile than the market and the other is less volatile than the market, you'll do better investing in the fund that is more volatile than the market because you'll get more of a market beating boost to your peformance from the effects of dollar cost averaging.jkm929
jrr7 you are asking good questions, and i'm not trying to be rhetorical by asking some back.so, an extra year or two or 5 or 10 has no value to you? is that what you are saying? i agree that for most individual investors, who are investing for a specific goal (college for the kids, the european vacation, retirement etc) that the bumpiness of the road is less important than whether they meet their goal on time. and hence MPT is not very applicable. however, it is not clear to me why you think that a permanent impairment of capital can not be defined for your set of circumstances.A 40% drop in the value of my portfolio this instant just means I need to work an additional year. It isn't permanent.that may well be true if your current capital base is small relative to your future capital investments and provided your time horizon is long. but it is still zero sum in the sense that the 40% of your capital that is impaired may have let you meet your goals two years early, or as you say you will wind up missing your goal by one year. again, does that time not have value?How does one know the probability of a large loss? that's a difficult one. small probability but high consequence events are very difficult to manage. i prefer self insurance (a portion of net worth in tbills) in that regard.if you do not understand the process that could impair your capital then it becomes difficult to define risk. this is what i consider the knightian uncertainty principle in its pure form.fwiw, i disagree that it was not possible to see the risks in LTCM or Enron beforehand. in both cases there was a level of leverage or senior claims on capital that almost by definition made the junior equity in both situations high risk.My point being that risk-adjusted returns are meaningless for me because they measure only volatility, which is an entirely different risk than the one I'm interested in.if risk=short run price volatility (90 day) that might be true for a long significant time horizon and provided you are not leveraged (no senior claims on your capital employed). not true if you are entrusting your capital to someone else.if risk=impairment of capital, then risk adjusted returns are important because impairment of capital risk is ubiquitous with all securities. for example, management at the business you are invested in could screw up and impair the company's capital, which in turn will impair the market price of your investment. inflation can seriously impair capital invested in a 10 year US treasury bond.tr
Okay. Now I think I'm beginning to understand what you're saying. From your POV (and which will be my POV pretty soon I hope), it's relevant.However, since my current capital is so small, and since the amount I am putting in is large compared to the capital value, don't the rewards of taking on additional risk, and the "dollar cost averaging" bonus of regularly investing in a volatile asset, more than make up for the probability of an impairment?Let 's put it this way. How can we calculate risk adjusted returns for impairment of capital when it's hard to manage the risks? Does the fundamentalist's benchmark "debt-to-equity ratio" figure into your calculations? How do you know when the mamangement isn't telling you the whole story? [Yes, I read the WSJ, but if you bail whenever they're nervous about a stock, you'll miss a lot of winners]BTW. I like TIPS in an IRA. I calculate the main risk as the government deciding wrong where to set the inflation numbers.
Best Of |
Favorites & Replies |
Start a New Board |
My Fool |