FTX had received several billion dollars in funding from major investors. One was a province pension fund, so it wasn’t just crypto folks. That generally involves having the investors’ accountants do substantial due diligence on the target firm’s financials. That tells me that either the books were fairly clean at the time of investment or they were cooked in a way that even the due dilligence specialists didn’t detect. It’s not clear to me how the Future Fund people, who to my knowledge are not forensic accountants or crypto experts, would have had a better ability to pick up on funny business. So I don’t see why it would be unreasonable for them to have relied on third-party expert vetting.
From what I understand (please correct me if I’m wrong), FTX didn’t have a CFO, it’s COO was a friend with no experience, and it didn’t have a proper board of directors. Clearly, that flimsy corporate governance would not pass a standard due diligence test.
EDIT: This flow chart of shells nested in shells, like Russian dolls, speaks to why the company’s governance should have been a red-flag.
In the autopsy, the biggest red flag will probably be the lack of appropriate internal controls. One should not be able to move that kind of money without vetting by staff with appropriate background and independence, but no ownership interest. Based on the reported en masse resignation of the bulk of legal and compliance staff, it seems that it was technically possible to transfer billions in customer assets to the CEO’s company without legal/compliance involvement.
I think the class of issues that would make it inappropriate to accept donations is much narrower than the issues that would and should make a public investor (like a province pension fund) decline to invest.
Few private businesses are going to let an outsider come in on a regular basis, conduct a hard look at sensitive internal documents, and potentially publish derogatory information to the public. Even for investors, this kind of stuff is generally done under a heavy NDA and for good reason. That would make it extremely difficult to do this on a regular basis—so any scrutiny would at best catch fraud that existed at the time of scrutiny.
This just isn’t plausible on reasonable priors. You need to assume that multiple investment firms working in different sectors, whose survival in a highly competitive environment in large part depends on being skilled at scrutinizing a company’s financials, would miss warning signs that should have been apparent to folks with no relevant domain expertise. See also Eliezer’s Twitter thread.
Some people are asking whether people who accepted FTX money should have “seen the red flags” or “done more due diligence”. Sometimes this is from outsider critics of effective altruism. More often it’s been effective altruists themselves, obsessively beating themselves up over dumb things like “I met an FTX employee once and he seemed to be frowning, why didn’t I realize that this meant they were all committing fraud?!” Listen: there’s a word for the activity of figuring out which financial entities are better or worse at their business than everyone else thinks, maximizing your exposure to the good ones, and minimizing your exposure to the bad ones. That word is “finance”. If you think you’re better at it than all the VCs, billionaires, and traders who trusted FTX—and better than all the competitors and hostile media outlets who tried to attack FTX on unrelated things while missing the actual disaster lurking below the surface—then please start a company, make $10 billion, and donate it to the victims of the last group of EAs who thought they were better at finance than everyone else in the world. Otherwise, please chill.
I would disagree, there are numerous examples such as Theranos and WeWork which show that sophisticated investors do not necessarily scrutinize potential investments thoroughly. Thus I don’t think assuming they do is a good prior. I think this is actually a reason these problems happen, since everyone else assumes that Respectable Company/Person X has scrutinized it.
I agree with the point that in general one should expect less from “unsophisticated” investors/parties than from sophisticated ones. I do not disagree with that.
I was disagreeing with “This just isn’t plausible on reasonable priors.” which seemed to mean that you disagreed with Stuart’s comment.
But I also don’t think VC scrutiny is necessarily a high bar in general in the absolute sense, and Stuart has posted some warning signs here in other comments such as the hiring of Friedberg. Then considering how important FTX and SBF was to the EA community it could have been investigated more, i.e. the low VC scrutiny bar could have been surpassed by hiring experts or something similar. To a VC firm this is just another losing bet among many they expect to make. This is why I don’t think the comparison with VC firms is very apt.
I was disagreeing with “This just isn’t plausible on reasonable priors.” which seemed to mean that you disagreed with Stuart’s comment.
Stuart’s comment was in reply to the claim that “It’s not clear to me how the Future Fund people, who to my knowledge are not forensic accountants or crypto experts, would have had a better ability to pick up on funny business.” I disagreed with Stuart’s comment in the sense that I disputed the reasonableness of expecting unsophisticated outsiders to do better because sophisticated investors sometimes perform poorly. I did not mean to dispute that sophisticated investors sometimes perform poorly; indeed, there’s plenty of evidence of that, including the evidence you provide in your comment.
FTX had received several billion dollars in funding from major investors. One was a province pension fund, so it wasn’t just crypto folks. That generally involves having the investors’ accountants do substantial due diligence on the target firm’s financials. That tells me that either the books were fairly clean at the time of investment or they were cooked in a way that even the due dilligence specialists didn’t detect. It’s not clear to me how the Future Fund people, who to my knowledge are not forensic accountants or crypto experts, would have had a better ability to pick up on funny business. So I don’t see why it would be unreasonable for them to have relied on third-party expert vetting.
From what I understand (please correct me if I’m wrong), FTX didn’t have a CFO, it’s COO was a friend with no experience, and it didn’t have a proper board of directors. Clearly, that flimsy corporate governance would not pass a standard due diligence test.
EDIT: This flow chart of shells nested in shells, like Russian dolls, speaks to why the company’s governance should have been a red-flag.
https://i.redd.it/078p4g7m6cz91.jpg
I don’t think a highly branched company structure is a red flag: my understanding is that to operate a financial business legally across many jurisdictions you generally need to have subsidiaries in each jurisdiction. Ex: https://wise.com/help/articles/2974131/what-are-the-wise-group-entities
In the autopsy, the biggest red flag will probably be the lack of appropriate internal controls. One should not be able to move that kind of money without vetting by staff with appropriate background and independence, but no ownership interest. Based on the reported en masse resignation of the bulk of legal and compliance staff, it seems that it was technically possible to transfer billions in customer assets to the CEO’s company without legal/compliance involvement.
I think the class of issues that would make it inappropriate to accept donations is much narrower than the issues that would and should make a public investor (like a province pension fund) decline to invest.
Few private businesses are going to let an outsider come in on a regular basis, conduct a hard look at sensitive internal documents, and potentially publish derogatory information to the public. Even for investors, this kind of stuff is generally done under a heavy NDA and for good reason. That would make it extremely difficult to do this on a regular basis—so any scrutiny would at best catch fraud that existed at the time of scrutiny.
I wouldn’t be very confident in the level of due diligence undertaken by supposedly sophisticated investors:
https://twitter.com/zebulgar/status/1590394857474109441
This just isn’t plausible on reasonable priors. You need to assume that multiple investment firms working in different sectors, whose survival in a highly competitive environment in large part depends on being skilled at scrutinizing a company’s financials, would miss warning signs that should have been apparent to folks with no relevant domain expertise. See also Eliezer’s Twitter thread.
ETA: Alexander:
I would disagree, there are numerous examples such as Theranos and WeWork which show that sophisticated investors do not necessarily scrutinize potential investments thoroughly. Thus I don’t think assuming they do is a good prior. I think this is actually a reason these problems happen, since everyone else assumes that Respectable Company/Person X has scrutinized it.
I am making a comparative, not an absolute, claim: however bad the professionals may be, it is unreasonable to expect outsiders to do better.
I agree with the point that in general one should expect less from “unsophisticated” investors/parties than from sophisticated ones. I do not disagree with that.
I was disagreeing with “This just isn’t plausible on reasonable priors.” which seemed to mean that you disagreed with Stuart’s comment.
But I also don’t think VC scrutiny is necessarily a high bar in general in the absolute sense, and Stuart has posted some warning signs here in other comments such as the hiring of Friedberg. Then considering how important FTX and SBF was to the EA community it could have been investigated more, i.e. the low VC scrutiny bar could have been surpassed by hiring experts or something similar. To a VC firm this is just another losing bet among many they expect to make. This is why I don’t think the comparison with VC firms is very apt.
Stuart’s comment was in reply to the claim that “It’s not clear to me how the Future Fund people, who to my knowledge are not forensic accountants or crypto experts, would have had a better ability to pick up on funny business.” I disagreed with Stuart’s comment in the sense that I disputed the reasonableness of expecting unsophisticated outsiders to do better because sophisticated investors sometimes perform poorly. I did not mean to dispute that sophisticated investors sometimes perform poorly; indeed, there’s plenty of evidence of that, including the evidence you provide in your comment.
Yeah that makes sense, I think I overinterpreted your comments.
In retrospect, I think my original comment was insufficiently clear. Anyway, thanks for the dialogue.
And that’s what Sequoia proudly publicly posted themselves
Or the best auditors are inadequate, and overlooked fairly obvious flaws for some reason.