4 ms·

They say later on that some firms do beat their benchmark over a 20 year period, so I don't think it's just noise.

maxbond 4y ago
You mean this part? > Some actively managed funds did better than the overall market over the last 15 or 20 years. Though they were unable to do so consistently year after year, they had good stretches, and those periods were strong enough to make them outperform over the entire span. Such funds may well be worth owning. > “Those that have managed to do that are impressive,” Mr. Edwards said. “But which funds will be able to do it over the next 20 years?” Unfortunately, we don’t know. If you owned that fund for 20 years you'd beat the benchmark. If you missed a few crucial moments - you wouldn't. It's really easy to get into the fund after a good year and leave after a bad year. The risk required for those big gains also sometimes results in big losses. Other times funds cheat, like Renaissance. They certainly had lots of quantitative innovations, but a huge part of their advantage was not paying their taxes. They settled with the IRS for 7 billion dollars. For the record, I don't think literally all funds doing well is noise, but the evidence seems to consistently bear out that most of the time it is.
- SideQuark 4y ago
  >but a huge part of their advantage was not paying their taxes. They settled with the IRS for 7 billion dollars. And that 7B is past taxes, interest, and penalties. They returned 66% annualized before fees and 39% after fees over a 30 year span, 1988 to 2018. Renaissance also has grown to over $130B. It also was not tied to many of their funds - it was only from (if I recall) a single fund. The others are not under IRS investigation (AFAIK) So the 7B is no where near enough of a cheat to allow this kind of return. You can view the 7B as evidence that the fun returned incredible returns to investors, so much so, that missed taxes on the profits were 7B.
  - maxbond 4y ago
    It was a $7B settlement, not that they failed to pay $7B dollars in taxes. Presumably they failed to pay much more than that. I don't know how much, I doubt it would be possible to calculate. But please correct me if this is known. ETA: > Levin in 2014 had presented the findings of a year-long probe into basket options, calling for tougher action from the authorities. The report said the largest user of the options, Renaissance Technologies Corp, saved an estimated $6.8 billion in taxes. https://www.reuters.com/business/finance/renaissance-executives-pay-about-7-bln-settle-tax-probe-wsj-2021-09-02/ https://www.reuters.com/business/finance/renaissance-executi... http://web.archive.org/web/20211226110616/https://www.reuters.com/business/finance/renaissance-executives-pay-about-7-bln-settle-tax-probe-wsj-2021-09-02/ http://web.archive.org/web/20211226110616/https://www.reuter... I kinda doubt that this could be accurately assessed by Levin (since the entire point was to add smoke and mirrors to thr ese transactions), but I'll concede that the best available evidence (at least after casual searching) suggests it was about $7B (in 2014). I don't really doubt that there was something to Renaissance's magic, but going by your numbers, nearly half of that magic was tax fraud. If your competitors are paying multiples more than you in taxes, because they're paying short term capital gains and you only pay long term - yeah, having the highest returns seems pretty feasible. I'd also note that "one fund" was their main fund, not like a little side project or something. Lastly I'd like to point out this linear, it's not like they would have returned 39% in a universe where they paid that $7B as that went instead of as a lump sum at the end. They basically were drawing on a line of credit from the taxpayer. They had additional liquidity and so they made additional money. You can't separate their tax fraud from their returns and say that their returns were just so good that they had a huge tax bill. No, they committed tax fraud, in order to get those returns.
    - SideQuark 4y ago
      >Presumably they failed to pay much more than that. I don't know how much, I doubt it would be possible to calculate. The exact amount is what the IRS figures out before damages are applied. >I kinda doubt that this could be accurately assessed by Levin Once someone starts to believe things that suit them without evidence they generally stop listening to actual evidence. This is exactly that pattern: an investigate, and likely the IRS also, did do an investigation likely of hundreds to thousands of man hours, reached the above conclusion, and you, an internet poster googling for a few minutes, disbelieves them? Financial News London, a pretty solid source on such things states "Current and former executives of hedge fund Renaissance Technologies will personally pay as much as $7bn in back taxes, interest and penalties to settle a long-running dispute with the Internal Revenue Service, the firm said, a tax settlement that may be the largest in history." [1] I think I'll trust their (and similar financial research places) on the characterization of the payment, especially since the standard for back payment has always been back taxes + penalties + interest. The IRS doesn't just play around, especially when something this high profile is at stake. >nearly half of that magic was tax fraud. What? Provide your napkin math. I just presented numbers that show this "half" to be far out of reality. If you're a programmer, simply set up a simple model (I just did that) and compute expected with or without paying that 7B over the past 30 years. The returns are vastly more than "half". Did you just claim half from any numerical checking, or just guessed it? >Lastly I'd like to point out this linear, it's not like they would have returned 39% in a universe where they paid that $7 The naive compound growth is T=P(1+r)^n, which is exactly linear. Double the principal P, double the outcome. Taxes aren't paid on value, only on profit, and if you check the math (it's pretty easy), taking X % each year or at the end of profits is exactly the same number. The practical reason profits are taxes yearly to to provide smoother income for govt and to get some of the profit before things tank and there is none left. In real investment, this actually becomes sublinear, as the more invested in a market, the lower percent returns, since the ability to grow is an S-curve (as is all things), not an exponential. Buffet, for example, has talked exteneively about this. This sublinear fact is also why Renaissance famously kicked out lots of early external investors - the funds under management were not able to grow since there was too much for the opportunities they chased. So I do agree it's not linear - it's sublinear. [1] https://www.fnlondon.com/articles/hedge-fund-renaissance-technologies-to-pay-7bn-to-settle-back-taxes-20210903 https://www.fnlondon.com/articles/hedge-fund-renaissance-tec...

You mean this part? > Some actively managed funds did better than the overall market over the last 15 or 20 years. Though they were unable to do so consistently year after year, they had good stretches, and those periods were strong enough to make them outperform over the entire span. Such funds may well be worth owning. > “Those that have managed to do that are impressive,” Mr. Edwards said. “But which funds will be able to do it over the next 20 years?” Unfortunately, we don’t know. If you owned that fund for 20 years you'd beat the benchmark. If you missed a few crucial moments - you wouldn't. It's really easy to get into the fund after a good year and leave after a bad year. The risk required for those big gains also sometimes results in big losses. Other times funds cheat, like Renaissance. They certainly had lots of quantitative innovations, but a huge part of their advantage was not paying their taxes. They settled with the IRS for 7 billion dollars. For the record, I don't think literally all funds doing well is noise, but the evidence seems to consistently bear out that most of the time it is.

SideQuark 4y ago
>but a huge part of their advantage was not paying their taxes. They settled with the IRS for 7 billion dollars. And that 7B is past taxes, interest, and penalties. They returned 66% annualized before fees and 39% after fees over a 30 year span, 1988 to 2018. Renaissance also has grown to over $130B. It also was not tied to many of their funds - it was only from (if I recall) a single fund. The others are not under IRS investigation (AFAIK) So the 7B is no where near enough of a cheat to allow this kind of return. You can view the 7B as evidence that the fun returned incredible returns to investors, so much so, that missed taxes on the profits were 7B.
- maxbond 4y ago
  It was a $7B settlement, not that they failed to pay $7B dollars in taxes. Presumably they failed to pay much more than that. I don't know how much, I doubt it would be possible to calculate. But please correct me if this is known. ETA: > Levin in 2014 had presented the findings of a year-long probe into basket options, calling for tougher action from the authorities. The report said the largest user of the options, Renaissance Technologies Corp, saved an estimated $6.8 billion in taxes. https://www.reuters.com/business/finance/renaissance-executives-pay-about-7-bln-settle-tax-probe-wsj-2021-09-02/ https://www.reuters.com/business/finance/renaissance-executi... http://web.archive.org/web/20211226110616/https://www.reuters.com/business/finance/renaissance-executives-pay-about-7-bln-settle-tax-probe-wsj-2021-09-02/ http://web.archive.org/web/20211226110616/https://www.reuter... I kinda doubt that this could be accurately assessed by Levin (since the entire point was to add smoke and mirrors to thr ese transactions), but I'll concede that the best available evidence (at least after casual searching) suggests it was about $7B (in 2014). I don't really doubt that there was something to Renaissance's magic, but going by your numbers, nearly half of that magic was tax fraud. If your competitors are paying multiples more than you in taxes, because they're paying short term capital gains and you only pay long term - yeah, having the highest returns seems pretty feasible. I'd also note that "one fund" was their main fund, not like a little side project or something. Lastly I'd like to point out this linear, it's not like they would have returned 39% in a universe where they paid that $7B as that went instead of as a lump sum at the end. They basically were drawing on a line of credit from the taxpayer. They had additional liquidity and so they made additional money. You can't separate their tax fraud from their returns and say that their returns were just so good that they had a huge tax bill. No, they committed tax fraud, in order to get those returns.
  - SideQuark 4y ago
    >Presumably they failed to pay much more than that. I don't know how much, I doubt it would be possible to calculate. The exact amount is what the IRS figures out before damages are applied. >I kinda doubt that this could be accurately assessed by Levin Once someone starts to believe things that suit them without evidence they generally stop listening to actual evidence. This is exactly that pattern: an investigate, and likely the IRS also, did do an investigation likely of hundreds to thousands of man hours, reached the above conclusion, and you, an internet poster googling for a few minutes, disbelieves them? Financial News London, a pretty solid source on such things states "Current and former executives of hedge fund Renaissance Technologies will personally pay as much as $7bn in back taxes, interest and penalties to settle a long-running dispute with the Internal Revenue Service, the firm said, a tax settlement that may be the largest in history." [1] I think I'll trust their (and similar financial research places) on the characterization of the payment, especially since the standard for back payment has always been back taxes + penalties + interest. The IRS doesn't just play around, especially when something this high profile is at stake. >nearly half of that magic was tax fraud. What? Provide your napkin math. I just presented numbers that show this "half" to be far out of reality. If you're a programmer, simply set up a simple model (I just did that) and compute expected with or without paying that 7B over the past 30 years. The returns are vastly more than "half". Did you just claim half from any numerical checking, or just guessed it? >Lastly I'd like to point out this linear, it's not like they would have returned 39% in a universe where they paid that $7 The naive compound growth is T=P(1+r)^n, which is exactly linear. Double the principal P, double the outcome. Taxes aren't paid on value, only on profit, and if you check the math (it's pretty easy), taking X % each year or at the end of profits is exactly the same number. The practical reason profits are taxes yearly to to provide smoother income for govt and to get some of the profit before things tank and there is none left. In real investment, this actually becomes sublinear, as the more invested in a market, the lower percent returns, since the ability to grow is an S-curve (as is all things), not an exponential. Buffet, for example, has talked exteneively about this. This sublinear fact is also why Renaissance famously kicked out lots of early external investors - the funds under management were not able to grow since there was too much for the opportunities they chased. So I do agree it's not linear - it's sublinear. [1] https://www.fnlondon.com/articles/hedge-fund-renaissance-technologies-to-pay-7bn-to-settle-back-taxes-20210903 https://www.fnlondon.com/articles/hedge-fund-renaissance-tec...