3 ms·

This is a useful insight, but with caveats. As an investor, my livelihood depends on my ability to consistently make sound decisions. And fortunately, decision

by dharmon 10y ago

This is a useful insight, but with caveats. As an investor, my livelihood depends on my ability to consistently make sound decisions. And fortunately, decisions are not set in stone, so I have opportunities to negate poor ones. I have thought a lot about how to improve my decision-making. One is what you are talking about, sharing my decisions and thought processes with others. An obvious conclusion (to me) is that doing so on the internet is horrible, as there is no real ability to filter opinions on my part, with no background knowledge of those giving feedback. Yet they can still sow seeds of self-doubt. But even talking with friends can be tricky, as you have to account for their backgrounds and biases. Sometimes it's easy. For example, if I'm looking at real estate, something I know little about, I would heavily count the thoughts of my friend who is a REIT analyst. Other times it is not clear how much to weight their opinion. A Bayesian approach seems to work here. Another idea which I find as effective, if not more, is constantly turning the decision over in my own mind. I am not the same person I was yesterday. My thought processes in the morning, after a cup of coffee, are different than Friday evening, after a few beers. It is definitely different after a period where the market is up ("I'm an invincible genius!") vs. down ("I'm a moron! Maybe I should go back to writing code...") By revisiting over a span of time, you can gather new insights solely based on where your head is at any given time. And at the end of the day, you have to honestly evaluate the results of your decisions. If I discount the opinions of others and follow my own intuition, but consistently lose money, then well, I should revisit some part of the process. Surprisingly many are unable (unwilling?) to do this.

qznc 10y ago
Kahneman in "Thinking Fast and Slo" has a section on when experts can trust their intuitions. For example, fire fighters can. Investment does not fit the criteria because the payout is inconsistent. You sometimes win although your decision was bad and sometimes you lose although your decision was sound. Your intuition gets fucked up by the weird outcomes.
- barry-cotter 10y ago
  Does this imply expertise in investment doesn't exist? Because the payout in poker is inconsistent but expertise clearly exists. The people who end up at the top table are much more consistent than would be expected by chance. There is also a relatively clear skill ranking. If two people play casino poker and consistently profit but one plays at 1-2 tables and the other at 10-20 the second is definitely better.
  - tsunamifury 10y ago
    He lables them "wicked markets" and essentially describes them as any market where any winning strategy is almost instantly mitigated and results in a loss. So not every poker market is wicked, which is why you see good players always seeking tables of suckers.
  - dharmon 10y ago
    No, he means it "fucks with your intuition" in the same way as poker, but potentially more confusing because the probabilities are not known. For example, in poker you may call with pocket Kings and lose, yet with hindsight it is still clear that you made the right decision. The probabilities say so. In investing, even after you lose money, it can sometimes be quite difficult to determine if you made a mistake or not, since the probabilities are never known. For example, did Mohnish Pabrai and Guy Spier make a mistake with Horsehead Holdings? The only clear mistake I can see is Pabrai's position sizing, but it is not clear to me whether the investment itself was a mistake or not.
  - qznc 10y ago
    There are poker experts, but they do not use their intuition (very much). Poker teaches you to make rational decisions while your intuition screams something different.
derefr 10y ago
> A Bayesian approach seems to work here. A tangent: is there a name for an automated investment approach that takes a large pool of investors, and tries to use Bayesian clustering to extract the "super-predictors" from the data, to base its investments on them?
arjie 10y ago
One thing I've been doing is limiting my 'true discussions'¹ to ones where I can reasonably say that both parties are acting in the specific discussion like they would approximate the conditions required for Aumann's agreement theorem². What follows is that every disagreement should be followed by new information. At the point where disagreements are close to "agree to disagree", I know that at least one of the two participants (either me or the other person) is not participating in an honest rational manner allowing me to note that the conversation should not be a means to inform myself. This is not an indictment of character. It's merely the nature of our minds that we are not always capable of usefully transmitting information to others. Repeating the discussion with a sufficient number of people will allow me to consider whether I'm the poor partner or whether it's the other people. Of course, the practice of this is somewhat removed from the description above, but I believe it's improved my ability to truth-find (judging by the degree to which my new opinions match reality) to even attempt this approach, even in a flawed manner. I _am_ doing better on the market now, but naturally it's impossible to tell if that's because I'm more well-informed or if that's luck. ¹ discussions where I'm attempting to actually expand my knowledge, as opposed to ones I'm participating in for social reasons ² people familiar with this will argue that you can't find people with the same priors as you, but is this really true? Shouldn't all rational agents arrive fully formed into the world void of any knowledge? If so, all you have to accept is fully conditioning everyone's statements on their entire experience from birth. Fortunately, this shouldn't be terribly hard if you both otherwise meet these conditions: http://www.scottaaronson.com/papers/agree-econ.pdf http://www.scottaaronson.com/papers/agree-econ.pdf
qwrusz 10y ago
I dig this comment. I think it speaks to the idea that when it comes to improving one's decision-making skills it's important to share your opinions / decisions with others when possible and also unabashedly share how you reached your opinion Unabashed openness in how and why you think something appears to be the hardest part for people. Maybe because it includes saying things like "I don't know" or saying to a group of people a sentence like "My opinion on xyz is based on intuition, I apologize I can't point to a more scientific reason for my opinion right now but just because something is a gut feeling doesn't mean it is incorrect". This kind of talk is especially difficult in corporate environments where, in general, managers or aspiring leaders saying something like "I don't know" is verboten. Q: You mentioned you are an investor. It wasn't clear what type of investing you do or if you do this professionally or if you're a one-man shop. But you also said something about getting feedback on decisions on the internet? (I also saw you said it is horrible :). I'm just curious where on the internet could you even have tried this getting feedback as an investor? While I don't know of a professional investment firm that would allow employees to discuss live investment decisions on the internet assuming it happens anyway; is there a website or something where professional investors discuss investment decisions and their theses online? I work in finance and invest professionally (I don't call myself an investor though maybe because most of the money belongs to other people). Anyway my experience has been (both at my current fund and previous fund) the PM decision-making processes and managing risks like bias or miscalculation are taken so so seriously now: There is a formal internal procedure for decisions that lead to investment portfolio changes and an external company/consulting firm is hired specifically to evaluate and improve decision making by humans and machines. Decision making has fallen into the "risk management" category and it is not taken lightly.
- dharmon 10y ago
  Sometimes my wife and I will go for walks and I talk her through my thinking on whatever I'm currently looking at. Even though she's not a finance person, just sharing my thought process with a super smart person helps me identify flaws or develop insights. I'm a one-man shop investing my own money (in plain jane public markets), so I have no restrictions on discussing. Value Investor's Club is good, as is Corner of Berkshire and Fairfax. I use Seeking Alpha because these people are amazing at digging up data. I just ignore the analysis cause it's usually pretty weak. I see people share their theses on these sites and places like Reddit and from the inconsistent quality in feedback they get I decided it would not be a valuable thing for me to do. To each their own, I guess. Crazy how institutionalized risk management is (hopefully it is beyond just VAR these days). My risk management is a little folksy by comparison: what's the most I could lose here? how likely is that to happen? and how much am I willing to lose? Are you an analyst at a fund?

by dharmon 10y ago

qznc 10y ago

Kahneman in "Thinking Fast and Slo" has a section on when experts can trust their intuitions. For example, fire fighters can. Investment does not fit the criteria because the payout is inconsistent. You sometimes win although your decision was bad and sometimes you lose although your decision was sound. Your intuition gets fucked up by the weird outcomes.

barry-cotter 10y ago
Does this imply expertise in investment doesn't exist? Because the payout in poker is inconsistent but expertise clearly exists. The people who end up at the top table are much more consistent than would be expected by chance. There is also a relatively clear skill ranking. If two people play casino poker and consistently profit but one plays at 1-2 tables and the other at 10-20 the second is definitely better.
- tsunamifury 10y ago
  He lables them "wicked markets" and essentially describes them as any market where any winning strategy is almost instantly mitigated and results in a loss. So not every poker market is wicked, which is why you see good players always seeking tables of suckers.
- dharmon 10y ago
  No, he means it "fucks with your intuition" in the same way as poker, but potentially more confusing because the probabilities are not known. For example, in poker you may call with pocket Kings and lose, yet with hindsight it is still clear that you made the right decision. The probabilities say so. In investing, even after you lose money, it can sometimes be quite difficult to determine if you made a mistake or not, since the probabilities are never known. For example, did Mohnish Pabrai and Guy Spier make a mistake with Horsehead Holdings? The only clear mistake I can see is Pabrai's position sizing, but it is not clear to me whether the investment itself was a mistake or not.
- qznc 10y ago
  There are poker experts, but they do not use their intuition (very much). Poker teaches you to make rational decisions while your intuition screams something different.

derefr 10y ago

> A Bayesian approach seems to work here. A tangent: is there a name for an automated investment approach that takes a large pool of investors, and tries to use Bayesian clustering to extract the "super-predictors" from the data, to base its investments on them?

arjie 10y ago

One thing I've been doing is limiting my 'true discussions'¹ to ones where I can reasonably say that both parties are acting in the specific discussion like they would approximate the conditions required for Aumann's agreement theorem². What follows is that every disagreement should be followed by new information. At the point where disagreements are close to "agree to disagree", I know that at least one of the two participants (either me or the other person) is not participating in an honest rational manner allowing me to note that the conversation should not be a means to inform myself. This is not an indictment of character. It's merely the nature of our minds that we are not always capable of usefully transmitting information to others. Repeating the discussion with a sufficient number of people will allow me to consider whether I'm the poor partner or whether it's the other people. Of course, the practice of this is somewhat removed from the description above, but I believe it's improved my ability to truth-find (judging by the degree to which my new opinions match reality) to even attempt this approach, even in a flawed manner. I _am_ doing better on the market now, but naturally it's impossible to tell if that's because I'm more well-informed or if that's luck. ¹ discussions where I'm attempting to actually expand my knowledge, as opposed to ones I'm participating in for social reasons ² people familiar with this will argue that you can't find people with the same priors as you, but is this really true? Shouldn't all rational agents arrive fully formed into the world void of any knowledge? If so, all you have to accept is fully conditioning everyone's statements on their entire experience from birth. Fortunately, this shouldn't be terribly hard if you both otherwise meet these conditions: http://www.scottaaronson.com/papers/agree-econ.pdf http://www.scottaaronson.com/papers/agree-econ.pdf

qwrusz 10y ago

I dig this comment. I think it speaks to the idea that when it comes to improving one's decision-making skills it's important to share your opinions / decisions with others when possible and also unabashedly share how you reached your opinion Unabashed openness in how and why you think something appears to be the hardest part for people. Maybe because it includes saying things like "I don't know" or saying to a group of people a sentence like "My opinion on xyz is based on intuition, I apologize I can't point to a more scientific reason for my opinion right now but just because something is a gut feeling doesn't mean it is incorrect". This kind of talk is especially difficult in corporate environments where, in general, managers or aspiring leaders saying something like "I don't know" is verboten. Q: You mentioned you are an investor. It wasn't clear what type of investing you do or if you do this professionally or if you're a one-man shop. But you also said something about getting feedback on decisions on the internet? (I also saw you said it is horrible :). I'm just curious where on the internet could you even have tried this getting feedback as an investor? While I don't know of a professional investment firm that would allow employees to discuss live investment decisions on the internet assuming it happens anyway; is there a website or something where professional investors discuss investment decisions and their theses online? I work in finance and invest professionally (I don't call myself an investor though maybe because most of the money belongs to other people). Anyway my experience has been (both at my current fund and previous fund) the PM decision-making processes and managing risks like bias or miscalculation are taken so so seriously now: There is a formal internal procedure for decisions that lead to investment portfolio changes and an external company/consulting firm is hired specifically to evaluate and improve decision making by humans and machines. Decision making has fallen into the "risk management" category and it is not taken lightly.

dharmon 10y ago
Sometimes my wife and I will go for walks and I talk her through my thinking on whatever I'm currently looking at. Even though she's not a finance person, just sharing my thought process with a super smart person helps me identify flaws or develop insights. I'm a one-man shop investing my own money (in plain jane public markets), so I have no restrictions on discussing. Value Investor's Club is good, as is Corner of Berkshire and Fairfax. I use Seeking Alpha because these people are amazing at digging up data. I just ignore the analysis cause it's usually pretty weak. I see people share their theses on these sites and places like Reddit and from the inconsistent quality in feedback they get I decided it would not be a valuable thing for me to do. To each their own, I guess. Crazy how institutionalized risk management is (hopefully it is beyond just VAR these days). My risk management is a little folksy by comparison: what's the most I could lose here? how likely is that to happen? and how much am I willing to lose? Are you an analyst at a fund?