Personally, I have a pretty strong independent streak, so the time-worn adage "Take in laundry before you take in partners" probably makes more sense to me than it does to others. But the whole point of partners (or, rather, their money, which you don't have ready to hand) is that they are effectively outsourcing their work ethic and judgement to you. In short, it's delayed gratification with interest.
In the long term, being a Luddite doesn't make sense. Then again, just because something's powered by machinery doesn't necessarily make it superior to what humans can do. Buying and selling the same stock within seconds--if, indeed, it actually takes that long--strikes me as one of those cases. The chiefest reason being that it doesn't make any sense from the standpoint of management...and partners are ultimately managers.
Admittedly, it's not an original management principle--in fact, I'm shamelessly stealing it wholesale from Joel Spolsky's Joel on Software blog. But the central idea is this: When you establish metrics, people (and in this case whole companies) will work toward those metrics--but the metrics are in fact only a sliver of what you actually want to accomplish. To illustrate: You want your call center to be more "productive," and you define that as the number of calls taken per caller per hour. So you establish rewards and punishments based upon the metric of calls/hour. Congratulations: You've just incentivized (to used management-speak) your phone-jock minions to hang up on your customers. Moron.
The one thing you need to understand about the buy/sell decisions of stock trades measured in microseconds is that computer programs--a.k.a. the algorithms--are just metrics. Ham fisted metrics. Why? Because algorithms can only model so much human behavior. Or, more to the point, so many possible outcomes. If you don't believe me, snag a Computer Science graduate and have her/him explain, say, the classic "Traveling Salesman" problem to you. It's an eye-opener regarding what algorithms can and can't do. Then throw in the tendency of options-compensated CEOs to cook the books, and you'll have as much faith in the base numbers as you should in sports records in the age of steroids.
So, in the meantime, I trust that you'll forgive me for not mistaking trades driven by algorithms as a functional substitute for actual investment guidance. Because so far as I'm concerned, that ain't capitalism. And, moreover, it's not something I'll gamble my golden years on.
