You are using an out of date browser. It may not display this or other websites correctly.
You should upgrade or use an alternative browser.
You should upgrade or use an alternative browser.
Trading Wisdom for Aspiring Hedge Fund Managers
- Thread starter darkhorse
- Start date
- Status
- Not open for further replies.
Quote from Ghost of Cutten:
You claimed that a losing streak is more dangerous for someone in a downswing than an upswing, even if they have identical present capital. Assuming a losing streak is the result of chance (rather than an edge disappearing, or a trader going on tilt), then the mortality risk is no greater for someone who went from 2 million to 1 million and is now betting 1% per trade, than from someone who went from 500k to 1 million and is now betting 1% per trade. If either trader has a huge losing streak, they will lose exactly the same amount, since they are making the same % bets, taking the same trades, and experiencing the same market outcomes.
I don't want to get into the semantics of one identical trader versus another, because it is a red herring relative to the point I was making.
If you always use a fixed fractional bet size of 1%, then yes, as your account size decreases in a losing streak the size of your bets will decrease.
But, if you are always fixed fractional 1%, then you are not utilizing the potential benefits of dynamic position sizing in the first place.
- If you always risk 1%, there is never a time when you risk 5% or more.
- Given changes in market conditions and conviction levels, there can be times when it is greatly advantageous to risk 5% or more.
- But if you risk 5% or more constantly, with no awareness of negative impact on equity curve in a losing streak, then you risk facing rapid decline.
Ask yourself these two questions:
In a rally car race, why do the best drivers vary their speeds at different points along the track? Why is it that, if one were to plot a histogram of average speeds for the winning car over the course of the race, a familiar bell curve shape would present itself, with outlier periods of slower than average (in bogs, rocky areas, hairpin turns etc) and also extremely fast (straightaways, hardpan flats etc?)?
Imagine if, in the same rally car race, one driver was given normal ability to speed up or slow down as he saw fit while the other was forced to drive a fixed average speed the whole time. Which driver would you rather bet on?
Quote from Ghost of Cutten:
Your other main claim was that what you call a 'profit cushion' has some kind of influence on results. This is pure mental accounting - accumulated profit is economically identical to initial capital: it buys the same things, funds trading positions exactly the same, it is indistinguishable. Otherwise brokers would allow reduced margin rates for 'accumulated P&L' rather than deposited funds, and luxury goods retailers would offer discounts to those whose dollars were earned by profitable trading, and charge more to people experiencing equity drawdowns. When Forbes calculate their rich list, they don't place a trader higher because he is at a new equity high, or mark someone lower because his stock fell 50% last year - they calculate based on how many dollars he has.
I still think you are processing it backwards. Accumulated profit reduces mortality risk, allowing for a wider range of speculative activity. This is common sense.
Why is it, for example, that Google is able to throw money at wacky projects like fiber-optic broadband in Kansas, 3D glasses, and self-driving cars - stuff that has an extremely high likelihood of failure (from a profit generating perspective) but a small chance of making tens of billions? Because they have accumulated enough profit to speculate at the margins of the business without jeopardizing the core of the business.
Similarly, why is a successful hedge fund able to expand into more areas of trading R&D, or look to more aggressive strategies in the market, than a fund that is just barely keeping its doors open? Because mortality risk - the risk of setback turning into disaster - is reduced with a cushion of profit.
Let us say you are presented with a trading opportunity in which you roughly estimate the odds of success as 1 out of 3 (33%), but the potential payoff at 10 to 1 or more. For every dollar you risk you stand to make ten dollars, but only with a 33% likelihood. How big would you size that trade?
Once again, cushion considerations matter. If you were up 35% for the year, you could put 5% into such a trade and, if you were wrong, c'est la vie but no worries. If you were down 5% on the year, however, risking an additional 5% of capital on a trade with a 2 out of 3 chance of failing would present too much danger to size aggressively.
As for all the stuff about brokers etc, that is red herring and irrelevant to the central point I am making.
Quote from Ghost of Cutten:
I agree about pro-cyclical market conditions (and other things like edges degrading, trading worse under psychological pressure etc), but one should adjust for these based on the relevant factors (present market conditions; present psychological state & sharpness of the trader), not the irrelevant ones (past drawdown, or past winning streak).
But P&L is actually a powerful signaling device in many cases. If you are consistently winning, it is a sign that your methodology is in synch with the market. If you are losing or getting chopped up, it is a sign or at least a possible strong indication that you are not in synch with the market.
This can be true even if your methodology is wholly mechanical. Take mechanical trend following futures traders, for example. The best of all environment for a mechanical trend following futures strategy would be a return to the steady inflation fear typified by the 1970s, in which commodity prices just went up and up and up, day after day. In an environment like that, what is a CTA trend follower's P&L going to look like? It is going to look fantastic.
Conversely, the worst of all macro environments for such a strategy - such as, say, when central bank activities are putting a collar on trend extremes up or down and leading to epic amounts of yo-yo chop n' slop - that will result in the P&L being put into a blender.
There are lots of ways to discuss this, but the bottom line point is that P&L itself can be a significant signal input in terms of how well you are trading and how conducive current conditions are to producing profits for your strategy, whether discretionary or mechanical. When the fitness landscape is trending P&L positive, you want more exposure; when the fitness landscape is trending P&L negative, you want less exposure.
Quote from Ghost of Cutten:
Winning streaks in poker have an effect on the plays of other players, in a way that doesn't occur in the markets, so it's not a good analogy. Aggressive betting is also superior in poker for reasons related to that card game's specifics; in the markets, aggressive betting is inferior for reasons that are specific to the nature of financial markets (i.e. unpredictable risks, edge-degredation, liquidity constraints, 'runs' by depositors or fund investors and so on). Furthermore, the mean-reverting nature of fund and star trader performance gives good reason to think the opposite - that you should become more sceptical and risk-averse after a long winning streak. Most experienced traders also know that winning streaks actually tend to lead to cockiness and over-betting (see the earlier Soros quote you gave), which is another reason not to press a hot hand.
Winning streaks absolutely have an impact on other players in markets too.
Consider, for example, what the composition of players looks like in the final stages of a major technology bull market. Who are going to be the biggest winners, the guys pumped full of testosterone and profits? They will be the ones who made the biggest, boldest bets on technology stocks. Those who were overly conservative will be left behind, and those who bet wrongly will be out of the game entirely.
Bull markets tend to end in spectacular flame-outs precisely because winning streaks have a large impact. The longer a strong bull market lasts, the more that the perceived "winners" are selected not for their prudence or risk control, but for the degree they were true believers and thus leveraged into the trend hugely.
The same applies in reverse in bear markets. By the end of a bear market, the majority of survivors are battle-scarred and cautious in the extreme. There is absolutely a pro-cyclical winning streak impact that is useful to consider in any thoughtful top down market analysis.
As for saying "aggressive betting is inferior" in markets - tell that to Stanley Druckenmiller and George Soros and Paul Tudor Jones!! And while you're at it tell it to Warren Buffett, who was just as aggressive as Soros in '92 or Jonesy in '87 or '90 in taking bet-of-a-lifetime positions in American Express, Washington Post, Geico, and Coca-Cola... if you really think that aggressive betting is inferior in markets, then you can argue with the track records of some of the biggest and best trading and investing legends ever to walk the earth.
As for not pressing a hot hand, you know who overpresses the hot hand? Bad traders or poorly skilled traders or otherwise less skilled than they think traders.
Knowing when to press a hot hand and when to be cautious and dial back is a matter of skill and methodological inputs; the fact that this requires wisdom, experience and talent is precisely why doing it right is a big source of alpha in the first place.
Quote from Brass:
I've read Rand's Atlas Shrugged and The Fountainhead several years ago. Pseudo-intellectual crap. She obviously had a hysterical knee-jerk reaction to the oppressive regime from which she hailed, and figured that if one extreme is destructive, then the other extreme must be the way to go. I think the attraction of her "philosophy" is two-fold. First it appeals to a lot of people's darker sides and validates their baser tendencies. Second, unlike other philospohies, it is much more accessible because it is so one-dimensional. Therefore, falling prey to false equivalency of philosophical thought, it is the lazy man's shortcut to becoming a "deep thinker."
Rand fascinates me precisely because she invites polarizing reactions such as this.
I personally think there is some real merit in Objectivism - see Victor Sperandeo's discussion of why he is an Objectivist in "Methods of a Wall Street Master" - but I also acknowledge the criticisms and sympathize with the viewpoint you more or less stated.
Quote from Ghost of Cutten:
But this assumes that i) rushes exist and are predictable ii) overbetting due to cockiness doesn't swamp the potential positive effects of rushes iii) there is no way to anticipate favourable trading conditions or performance other than witnessing prior abnormal profits.
Of course rushes exist. They are not predictable every time, but they are recognizable a meaningful percentage of the time, as are the market conditions that facilitate their birth and continuation.
"Overbetting due to cockiness" is what bad or inexperienced traders do. The skilled trader varies bet size based on keen observation and awareness of underlying dynamics shifting or sustaining in a favorable direction, and is quick to dial down when accumulated evidence (plus well honed intuition and market feel) tell him to do so.
Re, no way to anticipate favorable trading conditions, are you serious? Of course there is a way to anticipate favorable trading conditions, or to quickly recognize their presence not long after arrival - again not every time, but a meaningful percentage of the time.
You speak as someone who tautologically dismisses the argument because you believe the argument to be false. But there are plenty of inputs you overlook, in similar manner to the way an efficient market theorist might wave away the assertion that skill of any kind applies in any market at all.
Quote from Ghost of Cutten:
You are simply asserting/assuming that your view is correct, without providing any evidence to back it up.
Now you sound like one of those guys who are saying trends don't exist, or a Boglehead saying there isn't enough evidence traders actually have skill blah blah blah.
There is a lot of available evidence that what I am saying is true, a fair bit of it embedded in the track records and documented trading habits of the world's greatest traders and investors - and there is also an intrinsic logic to the case I am making that you seem to be dismissing out of hand. I also speak as a practitioner and not just a theoretician; I know the dynamics of this stuff intimately because I live and breathe the practice as well as the theory.
As for going out and digging up such evidence and making a legal presentation as if you were someone I needed to convince in a court of law beyond a shadow of reasonable doubt, no thanks - beyond a certain point you either get what I'm saying or you don't. There is no white paper being submitted for scholarly approval here.
I read both of Sperandeo's trading books in the '90s, and have no desire to revisit them. Particularly his Trader Vic II which, if memory serves, was perhaps more about his politics than his trading. TMI. Perhaps it comes as no surprise that his novel Crashmaker (vanity press?) was probably not much of a cashmaker.Quote from darkhorse:
Rand fascinates me precisely because she invites polarizing reactions such as this.
I personally think there is some real merit in Objectivism - see Victor Sperandeo's discussion of why he is an Objectivist in "Methods of a Wall Street Master" - but I also acknowledge the criticisms and sympathize with the viewpoint you more or less stated.
doesn't it make sense to exponentially decrease your position size at a faster rate when losing then when winning.. . 0 is always closer to infinity right?
i mean i can't afford to lose till i'm negative numbers... i have asymmetric risk .. 0 is always closer then infinity.. why not position size such that you push money towards the skew... towards the positive... multiplying your posiition size as you go up.. and multiplying it down faster as you go down.. .. you have more money when you win.. and less when you lose.. .it just seems like simple math to me...
i mean i can't afford to lose till i'm negative numbers... i have asymmetric risk .. 0 is always closer then infinity.. why not position size such that you push money towards the skew... towards the positive... multiplying your posiition size as you go up.. and multiplying it down faster as you go down.. .. you have more money when you win.. and less when you lose.. .it just seems like simple math to me...
Quote from cdcaveman:
doesn't it make sense to exponentially decrease your position size at a faster rate when losing then when winning.. . 0 is always closer to infinity right?
i mean i can't afford to lose till i'm negative numbers... i have asymmetric risk .. 0 is always closer then infinity.. why not position size such that you push money towards the skew... towards the positive... multiplying your posiition size as you go up.. and multiplying it down faster as you go down.. .. you have more money when you win.. and less when you lose.. .it just seems like simple math to me...
Yep. This is why we maintain the ability to vary position size by a factor of 40 to 1 or more. The flip side of the risk reduction argument is that, when you are smaller you can always go bigger, and the quality of the trading environment itself has a trend factor that tends to persist.
You're right, it ain't rocket science... not so much the math, though, but the underlying assumptions and drivers that some folks apparently have trouble with.
Quote from Brass:
I read both of Sperandeo's trading books in the '90s, and have no desire to revisit them. Particularly his Trader Vic II which, if memory serves, was perhaps more about his politics than his trading. TMI. Perhaps it comes as no surprise that his novel Crashmaker (vanity press?) was probably not much of a cashmaker.
Crashmaker, heh... that brings back memories. What a doorstopper of a book (or two books I should say, as it comes in two volumes).
Have to admit I enjoyed it though. Read it over a long weekend at a little b&b in the mountains some six or seven years ago. The pace would have been infuriating under normal circumstances, but it was nice for a relaxed vacation.
As with Objectivism, I see value in the Austrian economics view (while remaining aloof from 'pure' ideologies of any kind).
- Status
- Not open for further replies.