Michael Jordan was the world’s best basketball player, and insisted on testing himself against baseball, where he failed. Herbert Hoover was one of the world’s best businessmen, and insisted on testing himself against politics, where he crashed and burned. We’re all inmates in prisons of different names. Most of us accept it and get on with our lives. Adams couldn’t stop rattling the bars.
Which only leaves the initial claim that “at least for me this puts a final nail in the coffin of EMH.”
This is a polite way of hinting that you might be a brilliant investing wizard with the power to beat the market. Honestly, after making such a beautiful trade—and my gosh it really was beautiful—whom amongst us could resist that temptation? Certainly not me. And anyway, it might even be true!
Yesterday was the 6-year anniversary of my entry into the “beautiful” trade referenced above. On 2/10/2020 I cashed out ~10% of my investment portfolio and put it into S&P 500 April puts, a little more than a week before markets started crashing from COVID-19. The position first lost ~40% due to the market continuing to go up during that week, then went up to a peak of 30-50x (going by memory) before going to 0, with a final return of ~10x (due to partial exits along the way). After that, I dove into the markets and essentially traded full time for a couple of years, then ramped down my time/effort when the markets became seemingly more efficient over time (perhaps due to COVID stimulus money being lost / used up by retail traders), and as my portfolio outgrew smaller opportunities. (In other words, it became too hard to buy or sell enough stock/options in smaller companies without affecting its price. It seems underappreciated or not much talked about how much harder outperforming the market becomes as one’s assets under management grows. Also this was almost entirely equities and options. I stayed away from trading bonds, crypto, or forex.)
Starting with no experience in active trading/investing (I was previously 100% in index funds), my portfolio has returned a total of ~9x over these 6 years. (So ~4.5x or ~350% after the initial doubling, vs 127% for S&P 500. Also this is a very rough estimate since my trades were scattered over many accounts and it’s hard to back out the effects of other incomes and expenses, e.g. taxes.)
Of course without providing or analyzing the trade log (to show how much risk I was taking) it’s still hard to rule out luck. And if it was skill I’m not sure how to explain it, except to say that I was doing a lot of trial and error (looking for apparent mispricings around various markets, trying various strategies, scaling up or down strategies based on what seemed to work), guided by intuition and some theoretical understanding of finance and markets. If I’m doing something that can’t be easily replicated by any equally smart person, I’m not sure what it is.
Collection of my investing-related LW posts, which recorded some of this journey:
Maybe interesting to note my other near misses (aside from capturing only a fraction of the 30-50x from the COVID puts): billions of $ from mining Bitcoin if I started when it was first released, 350x from investing in Anthropic which I turned down due to moral qualms. Also could have sold my weidai.com domain name for $500k, a >1000x return, at the peak of its valuation (which turned out to be a bubble because the Chinese online loan sector that bid for the name went bust).
The explanation here seems to be that in retrospect my intellectual interests were highly correlated with extremely high return investment opportunities, and I had enough awareness/agency to capture some (but only some) of these opportunities. But how to explain this, when most people with intellectual interests seem to lack one or both of these features?
Why am I writing about this?
Partly because I’m not sure what lessons/conclusions I should draw from these experiences.
Partly to establish a public record. If nobody does (I think at least a couple of other people in the rationalist community may have achieved comparable returns (ex-crypto) but aren’t talking about it for privacy) it gives people a misleading view of the world.
As Scott Alexander’s post suggests, achieving success in multiple domains is rare, and people, including me, presumably attempt it in part so they can show off if they do achieve it.
I think yes, given the following benefits, with the main costs being opportunity cost and risk of losing a bunch of money in an irrational way (e.g. couldn’t quit if I turned out to be a bad trader), I think. Am I missing anything or did you have something in mind when asking this?
physical and psychic benefits of having greater wealth/security
social benefits (within my immediate family who know about it, and now among LW)
calibration about how much to trust my own judgment on various things
it’s a relatively enjoyable activity (comparable to playing computer games, which ironically I can’t seem to find the motivation to play anymore)
some small chance of eventually turning the money into fraction of lightcone
I was thinking mostly along the lines of, it sounds like you made money, but not nearly as much money as you could have made if you had instead invested in or participated more directly in DL scaling (even excluding the Anthropic opportunity), when you didn’t particularly need any money and you don’t mention any major life improvements from it beyond the nebulous (and often purely positional/zero-sum), and in the mean time, you made little progress on past issues of importance to you like decision theory while not contributing to DL discourse or more exotic opportunities which were available 2020-2025 (like doing things like, eg. instill particular decision theories into LLMs by writing online during their most malleable years).
Thanks for clarifying! I was pretty curious where you were coming from.
not nearly as much money as you could have made if you had instead invested in or participated more directly in DL scaling (even excluding the Anthropic opportunity)
Seems like these would all have similar ethical issues as investing in Anthropic, given that I’m pessimistic about AI safety and want to see an AI pause/stop.
when you didn’t particularly need any money and you don’t mention any major life improvements from it beyond the nebulous
To be a bit more concrete, the additional wealth allowed us to escape the political dysfunction of our previous locality and move halfway across the country (to a nicer house/location/school) with almost no stress, and allows us not to worry about e.g. Trump craziness affecting us much personally since we can similarly buy our way out of most kinds of trouble (given some amount of warning).
(and often purely positional/zero-sum)
These are part of my moral parliament or provisional values. Do you think they shouldn’t be? (Or what is the relevance of pointing this out?)
you made little progress on past issues of importance to you like decision theory
By 2020 I had already moved away from decision theory and my new area of interest (metaphilosophy) doesn’t have an apparent attack so I mostly just kept it in the back of my mind as I did other things and waited for new insights to pop up. I don’t remember how I was spending my time before 2020, but looking at my LW post history, it looks like mostly worrying about wokeness, trying to find holes in Paul Christiano’s IDA, and engaging with AI safety research in general, none of which looks super high value in retrospect.
More generally I often give up or move away from previous interests (crypto and programming being other examples) and this seems to work for me.
eg. instill particular decision theories into LLMs by writing online during their most malleable years
Maybe to rule out luck you could give an estimate of your average beta and Sharpe ratio, which mostly only depend on your returns over time. Also, are you planning to keep actively trading part-time?
This seemed like a good idea that I spent some time looking into, but ran into a roadblock. My plan was to download all the monthly statements of my accounts (I verified that they’re still available, but total more than 1000 so would require some AI assistance/coding just to download/process), build a dataset of the monthly balances, then produce the final stats from the monthly total balances. But when I picked two consecutive monthly statements of a main account to look at, the account value decreased 20% from one month to the next and neither I nor the two AIs I asked (Gemini 3.0 Pro and Perplexity w/ GPT 5.2) could figure out why by looking at the 70+ page statement.[1] Eventually Gemini hallucinated an outgoing transfer as the explanation, and Perplexity claimed that it can’t give an answer because it doesn’t have a full list of positions (which is clearly in the statement that I uploaded to it). Maybe I’ll try to investigate some more in the future, but at this point it’s looking like more hassle than it’s worth.
I had a large position of SPX options, in part for box spread financing, and their values are often misreported when I look at them in my accounts online. But this doesn’t seem to explain the missing 20% in this case.
I was also redeeming SPACs for their cash value, which would cause the position and associated value to disappear from the account for like a week before coming back, which would require AI assistance to compensate for if I went through with the plan. But this doesn’t seem to explain the missing 20% for this month either.
perhaps due to COVID stimulus money being lost / used up by retail traders
If most people in the US had a bank account that featured monthly payments anywhere close to the “interest rate,” the government could reduce risky retail investments with little delay by raising rates. This is not the case. Assuming even highly bounded rationality, it seems like retail traders should still not be losing as much money as they do, so maybe I’m making a modeling mistake and it would turn out that people really dislike bank accounts. This may be a typical mind fallacy problem, but I have some evidence that’s not the case. Either way, it seems like when you distribute stimulus to consumers[1] and they make high variance (or downright stupid) moves, large portions of the stimulus money will end up doing something similar to an unbalanced version of Japan-like QE[2] after being taken from retail traders by automated bots. The central bank could try to correct the balance away from equities by increasing interest rates. Maybe the models are better today than they were 20 years ago, but retail-stupidity-rate seems hard to estimate in advance. It seems like you might get a QE-like effect at the wrong time.
I wonder if that had something to do with why there appeared to be such a large deviation from the EMH even after your first year of active trading. It seems fuzzy in my mind how the mechanics would work though, and I generally wonder why the trading bots didn’t do better against you. Was their working capital locked up elsewhere?
As an aside, AI systems that are persuasive but otherwise not especially competent could have major influence on what silly investments (or “investments”) people make. I wouldn’t even know where to start if I was presented with a lump-sum UBI proposal in a few years because of things like this. If human consumers can’t hang on to the majority of the sum for long enough, certain interest rates may start hitting legal limits, causing terrible distortion. Note that this runs through the QE-like-effect argument from before, and assumes the government is too slow/ineffective and can’t confiscate everything it can and (sometimes physically) burn everything it can’t. Confiscate-and-redistribute isn’t QE and destruction of the economy can limit intelligence-explosion like upwards effects on interest rates.
Do you regret not investing in Anthropic? I don’t know how much the investment was for, but it seems like you could do a lot of good with 350x that amount. Is there a return level you would have been willing to invest in it for (assuming the return level was inevitable; you would not be causing the company to increase by 1000x)?
I don’t regret it, and part of the reason is that I find it hard to find people/opportunities to direct resources to that I can be confident won’t end up doing more harm than good. Reasons:
Meta: Well-meaning people often end up making things worse. See Anthropic (and many other examples), and this post.
Object-level: It’s really hard to find people who share enough of my views that I can trust their strategy / decision making. For example when MIRI was trying to build FAI I thought they should be pushing for AI pause/stop, and now that they’re pushing for AI pause/stop, I worry they’re focusing too much on AI misalignment (to the exclusion of other similarly concerning AI-related risks) as well as being too confident in misalignment. I think this could cause a backlash in the unlikely (but not vanishingly so) worlds where AI alignment turns out to be relatively easy but we still need to solve other AI x-risks.
(When I did try to direct resources to others in the past, I often regretted it later. I think the overall effect is unclear or even net negative. Seems like it would have to be at least “clearly net positive” to justify investing in Anthropic as “earning-to-give”.)
How much do you think the skill you used is the basic superforcaster skill? Did you do Metaculus or GJOpen and think people with similar forcasting skills are likely also going to be good at investing or do you think you had different skill that go beyond that?
It seems like a good question, but unfortunately I have no familiarity with superforecasting, having never learned about it or participated in anything related except by reading some superficial descriptions of what it is.
Until Feb 2020 I had little interest in making empirical forecasts, since I didn’t see it as part of my intellectual interests, and believed in EMH or didn’t think it would be worth my time/effort to try to beat the market, so I just left such forecasting to others and deferred to other people who seem to have good epistemics.
If I had to guess based on my shallow understanding of superforecasting, I would say while there are probably overlapping skills, there’s a strategic component to trading, which involves things like which sectors to allocate attention to, how to spot the best opportunities and allocate capital to them, while not taking too much concentrated risk, explore vs exploit type decisions, which are not part of superforecasting.
If you’re still interested in trading(although maybe you’re not so much given the possibility of the impending singularity) maybe you should try polymarket, the returns there can be pretty good for smart people even if they have a lot of money. I 5x-ed in 2.5 years starting with four figures, but other people have done much better starting with a similar amount(up to 6 or 7 figures in a similar time frame), and my impression is that 2X for people with 7 figures should be achievable.
My 6 years as a trader / active investor
The Dilbert Afterlife by Scott Alexander, Jan 16, 2026:
The EMH Aten’t Dead by Richard Meadows, May 15, 2020:
Yesterday was the 6-year anniversary of my entry into the “beautiful” trade referenced above. On 2/10/2020 I cashed out ~10% of my investment portfolio and put it into S&P 500 April puts, a little more than a week before markets started crashing from COVID-19. The position first lost ~40% due to the market continuing to go up during that week, then went up to a peak of 30-50x (going by memory) before going to 0, with a final return of ~10x (due to partial exits along the way). After that, I dove into the markets and essentially traded full time for a couple of years, then ramped down my time/effort when the markets became seemingly more efficient over time (perhaps due to COVID stimulus money being lost / used up by retail traders), and as my portfolio outgrew smaller opportunities. (In other words, it became too hard to buy or sell enough stock/options in smaller companies without affecting its price. It seems underappreciated or not much talked about how much harder outperforming the market becomes as one’s assets under management grows. Also this was almost entirely equities and options. I stayed away from trading bonds, crypto, or forex.)
Starting with no experience in active trading/investing (I was previously 100% in index funds), my portfolio has returned a total of ~9x over these 6 years. (So ~4.5x or ~350% after the initial doubling, vs 127% for S&P 500. Also this is a very rough estimate since my trades were scattered over many accounts and it’s hard to back out the effects of other incomes and expenses, e.g. taxes.)
Of course without providing or analyzing the trade log (to show how much risk I was taking) it’s still hard to rule out luck. And if it was skill I’m not sure how to explain it, except to say that I was doing a lot of trial and error (looking for apparent mispricings around various markets, trying various strategies, scaling up or down strategies based on what seemed to work), guided by intuition and some theoretical understanding of finance and markets. If I’m doing something that can’t be easily replicated by any equally smart person, I’m not sure what it is.
Collection of my investing-related LW posts, which recorded some of this journey:
Look for the Next Tech Gold Rush?
Buying COVID puts
Tips/tricks/notes on optimizing investments
Anti-EMH Evidence (and a plea for help)
How to bet against civilizational adequacy?
Other near misses
Maybe interesting to note my other near misses (aside from capturing only a fraction of the 30-50x from the COVID puts): billions of $ from mining Bitcoin if I started when it was first released, 350x from investing in Anthropic which I turned down due to moral qualms. Also could have sold my weidai.com domain name for $500k, a >1000x return, at the peak of its valuation (which turned out to be a bubble because the Chinese online loan sector that bid for the name went bust).
The explanation here seems to be that in retrospect my intellectual interests were highly correlated with extremely high return investment opportunities, and I had enough awareness/agency to capture some (but only some) of these opportunities. But how to explain this, when most people with intellectual interests seem to lack one or both of these features?
Why am I writing about this?
Partly because I’m not sure what lessons/conclusions I should draw from these experiences.
Partly to establish a public record. If nobody does (I think at least a couple of other people in the rationalist community may have achieved comparable returns (ex-crypto) but aren’t talking about it for privacy) it gives people a misleading view of the world.
As Scott Alexander’s post suggests, achieving success in multiple domains is rare, and people, including me, presumably attempt it in part so they can show off if they do achieve it.
Ex post or ex ante, do you feel like this was ultimately a good use of your time starting in mid-2020? (I might have asked you this already.)
I think yes, given the following benefits, with the main costs being opportunity cost and risk of losing a bunch of money in an irrational way (e.g. couldn’t quit if I turned out to be a bad trader), I think. Am I missing anything or did you have something in mind when asking this?
physical and psychic benefits of having greater wealth/security
social benefits (within my immediate family who know about it, and now among LW)
calibration about how much to trust my own judgment on various things
it’s a relatively enjoyable activity (comparable to playing computer games, which ironically I can’t seem to find the motivation to play anymore)
some small chance of eventually turning the money into fraction of lightcone
evidence about whether I’m in a simulation
some marginal increase in credibility for my ideas
I was thinking mostly along the lines of, it sounds like you made money, but not nearly as much money as you could have made if you had instead invested in or participated more directly in DL scaling (even excluding the Anthropic opportunity), when you didn’t particularly need any money and you don’t mention any major life improvements from it beyond the nebulous (and often purely positional/zero-sum), and in the mean time, you made little progress on past issues of importance to you like decision theory while not contributing to DL discourse or more exotic opportunities which were available 2020-2025 (like doing things like, eg. instill particular decision theories into LLMs by writing online during their most malleable years).
Thanks for clarifying! I was pretty curious where you were coming from.
Seems like these would all have similar ethical issues as investing in Anthropic, given that I’m pessimistic about AI safety and want to see an AI pause/stop.
To be a bit more concrete, the additional wealth allowed us to escape the political dysfunction of our previous locality and move halfway across the country (to a nicer house/location/school) with almost no stress, and allows us not to worry about e.g. Trump craziness affecting us much personally since we can similarly buy our way out of most kinds of trouble (given some amount of warning).
These are part of my moral parliament or provisional values. Do you think they shouldn’t be? (Or what is the relevance of pointing this out?)
By 2020 I had already moved away from decision theory and my new area of interest (metaphilosophy) doesn’t have an apparent attack so I mostly just kept it in the back of my mind as I did other things and waited for new insights to pop up. I don’t remember how I was spending my time before 2020, but looking at my LW post history, it looks like mostly worrying about wokeness, trying to find holes in Paul Christiano’s IDA, and engaging with AI safety research in general, none of which looks super high value in retrospect.
More generally I often give up or move away from previous interests (crypto and programming being other examples) and this seems to work for me.
I would not endorse doing this.
Maybe to rule out luck you could give an estimate of your average beta and Sharpe ratio, which mostly only depend on your returns over time. Also, are you planning to keep actively trading part-time?
This seemed like a good idea that I spent some time looking into, but ran into a roadblock. My plan was to download all the monthly statements of my accounts (I verified that they’re still available, but total more than 1000 so would require some AI assistance/coding just to download/process), build a dataset of the monthly balances, then produce the final stats from the monthly total balances. But when I picked two consecutive monthly statements of a main account to look at, the account value decreased 20% from one month to the next and neither I nor the two AIs I asked (Gemini 3.0 Pro and Perplexity w/ GPT 5.2) could figure out why by looking at the 70+ page statement.[1] Eventually Gemini hallucinated an outgoing transfer as the explanation, and Perplexity claimed that it can’t give an answer because it doesn’t have a full list of positions (which is clearly in the statement that I uploaded to it). Maybe I’ll try to investigate some more in the future, but at this point it’s looking like more hassle than it’s worth.
I had a large position of SPX options, in part for box spread financing, and their values are often misreported when I look at them in my accounts online. But this doesn’t seem to explain the missing 20% in this case.
I was also redeeming SPACs for their cash value, which would cause the position and associated value to disappear from the account for like a week before coming back, which would require AI assistance to compensate for if I went through with the plan. But this doesn’t seem to explain the missing 20% for this month either.
Actual returns (so far) closer to 100x due to dilution. (Company has ~500x’d in valuation, but value of Series A shares has ~100x’d.)
If most people in the US had a bank account that featured monthly payments anywhere close to the “interest rate,” the government could reduce risky retail investments with little delay by raising rates. This is not the case. Assuming even highly bounded rationality, it seems like retail traders should still not be losing as much money as they do, so maybe I’m making a modeling mistake and it would turn out that people really dislike bank accounts. This may be a typical mind fallacy problem, but I have some evidence that’s not the case. Either way, it seems like when you distribute stimulus to consumers[1] and they make high variance (or downright stupid) moves, large portions of the stimulus money will end up doing something similar to an unbalanced version of Japan-like QE[2] after being taken from retail traders by automated bots. The central bank could try to correct the balance away from equities by increasing interest rates. Maybe the models are better today than they were 20 years ago, but retail-stupidity-rate seems hard to estimate in advance. It seems like you might get a QE-like effect at the wrong time.
I wonder if that had something to do with why there appeared to be such a large deviation from the EMH even after your first year of active trading. It seems fuzzy in my mind how the mechanics would work though, and I generally wonder why the trading bots didn’t do better against you. Was their working capital locked up elsewhere?
[3]
A more full analysis would involve some sort of model of PPP fraud, but I’m not sure how easy that would be.
https://en.wikipedia.org/wiki/Helicopter_money
https://en.wikipedia.org/wiki/Quantitative_easing
As an aside, AI systems that are persuasive but otherwise not especially competent could have major influence on what silly investments (or “investments”) people make. I wouldn’t even know where to start if I was presented with a lump-sum UBI proposal in a few years because of things like this. If human consumers can’t hang on to the majority of the sum for long enough, certain interest rates may start hitting legal limits, causing terrible distortion. Note that this runs through the QE-like-effect argument from before, and assumes the government is too slow/ineffective and can’t confiscate everything it can and (sometimes physically) burn everything it can’t. Confiscate-and-redistribute isn’t QE and destruction of the economy can limit intelligence-explosion like upwards effects on interest rates.
Do you regret not investing in Anthropic? I don’t know how much the investment was for, but it seems like you could do a lot of good with 350x that amount. Is there a return level you would have been willing to invest in it for (assuming the return level was inevitable; you would not be causing the company to increase by 1000x)?
I don’t regret it, and part of the reason is that I find it hard to find people/opportunities to direct resources to that I can be confident won’t end up doing more harm than good. Reasons:
Meta: Well-meaning people often end up making things worse. See Anthropic (and many other examples), and this post.
Object-level: It’s really hard to find people who share enough of my views that I can trust their strategy / decision making. For example when MIRI was trying to build FAI I thought they should be pushing for AI pause/stop, and now that they’re pushing for AI pause/stop, I worry they’re focusing too much on AI misalignment (to the exclusion of other similarly concerning AI-related risks) as well as being too confident in misalignment. I think this could cause a backlash in the unlikely (but not vanishingly so) worlds where AI alignment turns out to be relatively easy but we still need to solve other AI x-risks.
(When I did try to direct resources to others in the past, I often regretted it later. I think the overall effect is unclear or even net negative. Seems like it would have to be at least “clearly net positive” to justify investing in Anthropic as “earning-to-give”.)
How much do you think the skill you used is the basic superforcaster skill? Did you do Metaculus or GJOpen and think people with similar forcasting skills are likely also going to be good at investing or do you think you had different skill that go beyond that?
It seems like a good question, but unfortunately I have no familiarity with superforecasting, having never learned about it or participated in anything related except by reading some superficial descriptions of what it is.
Until Feb 2020 I had little interest in making empirical forecasts, since I didn’t see it as part of my intellectual interests, and believed in EMH or didn’t think it would be worth my time/effort to try to beat the market, so I just left such forecasting to others and deferred to other people who seem to have good epistemics.
If I had to guess based on my shallow understanding of superforecasting, I would say while there are probably overlapping skills, there’s a strategic component to trading, which involves things like which sectors to allocate attention to, how to spot the best opportunities and allocate capital to them, while not taking too much concentrated risk, explore vs exploit type decisions, which are not part of superforecasting.
If you’re still interested in trading(although maybe you’re not so much given the possibility of the impending singularity) maybe you should try polymarket, the returns there can be pretty good for smart people even if they have a lot of money. I 5x-ed in 2.5 years starting with four figures, but other people have done much better starting with a similar amount(up to 6 or 7 figures in a similar time frame), and my impression is that 2X for people with 7 figures should be achievable.