one of the things about my line of work (finance) is that i appropriate its jargon and way of thinking into my everyday life. this, of course, happened when i was a mathematician: i was constantly throwing around words like isomorphism, which is a perfectly fine non-math word but not really one that non-math people would have in their active vocabulary. (side note: the lexicon has outlived the profession.) now i'm constantly talking about arbitrage and (market-style) inefficiency and whatnot.

but there's an even more relevant analogy. as fate would have it, my first project for the firm was on risk management. the concept (i think i'm not breaking any sort of secrecy here; everything i'm telling you is pretty universal) is simple: you want to diversify your portfolio to minimize risk. suppose you have two stocks, both with 30 percent expected return and 30 percent risk (meaning that the return is a random variable with mean 30 (percent) and standard deviation 30.)

now, if you invest all of your wealth in one stock, you will have characteristics 30/30 on your portfolio. but if you invest half of your wealth in one stock and half in the other then, assuming they are uncorrelated, you will still have an expected return of 30, but your expected risk will now be about 21 percent. since most people's utility curves on wealth show diminishing marginal returns, this is a good thing.

okay, i don't blog about economics, so where am i going with this? i think one of my fundamental problems in life, as far as achieving the life that i want, is that i am a high-risk, high-reward stock. what this means is that i can be a tremendous asset as part of a diversified portfolio -- adding trace amounts of a stock with phenomenal return and extremely high risk does much more good to your portfolio return than bad to your risk (trust me on the stats, or do it for yourself, either way) -- but i'm not a great bet to sink your retirement into.

so what's the logical conclusion? great friend, lousy life partner, perhaps. i' m capable of being incredibly awesome to be around (awesome meaning smart, or funny, or cute, or whatever), but i'm also capable of being an entirely negative influence (depression spells, or general volatility, or high-maintenance, or clingy, or combative, or sleep-deprived and cranky: that's probably all of my main failure modes.)

to be fair, i think everyone perceives themselves as more moody than other people perceive them, but nonetheless i don't think my friends would argue with the statement that i am both more awesome and more mercurial than just about anyone else out there. which makes me a great part of a diversified portfolio, and probably helps explain why i have so many great friends. but the flip side is that i don't provide the steady growth of stability; there are going to be down days, down months, down years. and if i'm the primary part of your portfolio, this badness gets highlighted. and i think that's part of why, from a purely objective standpoint, relationship success is hard for me (or to put it more subjectively, why i'm bad at relationships.)

but there is good news for my kind. it turns out that a portfolio with 30 percent return and 30 percent risk, annually, has a 96 percent chance of outperforming a portfolio with 10 percent return and 5 percent risk, over a 10 year period, assuming no autocorrelation (perhaps a dubious assumption.) even a portfolio with 30 percent return and 50 percent risk versus 10/5 wins 88 percent of the time over a 30-year period (a conservative estimate of the amount of life-partnership.)

i feel certain that i have convinced my hypothetical robot girlfriend to stay with me.

back to the weblog