I feel like there is a difference between the distribution of VC returns and the distribution of S&P 500 component returns. A VC fund is sort of an equal weight enterprise - write 50 $10 million checks to early stage companies and expect one to return the fund. The S&P 500 is weighted by size, because it is constructed the experience of investors as a whole, and since the distribution of company size is skewed, the top 5 companies make up 20% of the index, the top 100 companies make up 75%, and so on, so even if every company performs equally, a handful of companies will account for the majority of returns - in fact as long as some companies go bankrupt or have negative returns, almost by definition a handful of companies will account for 100% of returns (you have the companies with negative returns, some companies that have just enough positive returns to offset the companies with negative returns, and the remainder therefore account for "100% of returns", in kind of a misleading way).
My point is, yes there is a wide distribution of component returns, but my guess is that the concentration of returns in the S&P is as much or more a function of component weighting as it is dispersion of returns.
Re TIPS, this reminds me of the original QE argument that the Treasury is picking which bonds to issue in the first place, they're big enough where could presumably drive down long term rates by issuing more 3 month T-bills and fewer 10 year notes instead of this whole roundabout method where the Fed buys longer term notes.
(Scott Sumner had an interesting point along the same lines that in the early 80s, the Treasury Department was just giving away billions of dollars by issuing 30 year bonds at 14% when Volcker was planning on crushing inflation. Why don't they coordinate? Or why isn't bond issuance just under the purview of the Fed?)
So in this case, how is the Treasury picking TIPS issuance vs 10 year issuance? If they issue more based on when the breakeven inflation rate goes up, then anything the Fed does to buy TIPS will be exactly offset by the Treasury issuing more.
If I had to guess, I would bet the Fed tries to stay neutral by buying TIPS and normal notes proportionally to the amounts they are issued, and so it's not really a factor.
Anyway, breakeven inflation implied by TIPS seems to be only a bit higher than usual, the main factor seems to be legitimately low real rates, which are somewhat globally determined based on savings and investment, and with the world where it is, that's not terribly surprising.
On the VC/weighting question, that misses follow-on rounds, which anecdotally seem to drive the really extraordinary returns, in dollar but not percentage terms. You're right that the S&P skewness is somewhat tautological, but it's a very good exercise for someone who is running a concentrated portfolio of companies that aren't a big piece of the index. They have to be very confident that they're making good picks.
>If I had to guess, I would bet the Fed tries to stay neutral by buying TIPS and normal notes proportionally to the amounts they are issued, and so it's not really a factor.
But they've bought more than 100% of net TIPS issuance since the pandemic hit! It's weird.
I guess the thought experiment with TIPS is, if the Treasury doubled or halved the supply of TIPS (relative to the overall supply of debt), would that really change the market breakeven inflation rate that much? And if the answer is yes, then why even bother issuing TIPS in the first place?
I'm not sure what the original goal is, but if it's to measure inflation expectations, then it's pointless if the price is that sensitive to supply (like if breakeven inflation is 3% if supply is doubled and 1% if it's halved), and if it's to lower the govt's interest expense, then if price is sensitive to supply, they should reduce supply to drive up the price until breakeven inflation is way too high, either by the Fed buying more or Treasury issuing less.
Thanks for the link to the Hamming talk, I didn't expect to read the entirely of it but ended up really enjoying it!
I feel like there is a difference between the distribution of VC returns and the distribution of S&P 500 component returns. A VC fund is sort of an equal weight enterprise - write 50 $10 million checks to early stage companies and expect one to return the fund. The S&P 500 is weighted by size, because it is constructed the experience of investors as a whole, and since the distribution of company size is skewed, the top 5 companies make up 20% of the index, the top 100 companies make up 75%, and so on, so even if every company performs equally, a handful of companies will account for the majority of returns - in fact as long as some companies go bankrupt or have negative returns, almost by definition a handful of companies will account for 100% of returns (you have the companies with negative returns, some companies that have just enough positive returns to offset the companies with negative returns, and the remainder therefore account for "100% of returns", in kind of a misleading way).
My point is, yes there is a wide distribution of component returns, but my guess is that the concentration of returns in the S&P is as much or more a function of component weighting as it is dispersion of returns.
Re TIPS, this reminds me of the original QE argument that the Treasury is picking which bonds to issue in the first place, they're big enough where could presumably drive down long term rates by issuing more 3 month T-bills and fewer 10 year notes instead of this whole roundabout method where the Fed buys longer term notes.
(Scott Sumner had an interesting point along the same lines that in the early 80s, the Treasury Department was just giving away billions of dollars by issuing 30 year bonds at 14% when Volcker was planning on crushing inflation. Why don't they coordinate? Or why isn't bond issuance just under the purview of the Fed?)
So in this case, how is the Treasury picking TIPS issuance vs 10 year issuance? If they issue more based on when the breakeven inflation rate goes up, then anything the Fed does to buy TIPS will be exactly offset by the Treasury issuing more.
If I had to guess, I would bet the Fed tries to stay neutral by buying TIPS and normal notes proportionally to the amounts they are issued, and so it's not really a factor.
Anyway, breakeven inflation implied by TIPS seems to be only a bit higher than usual, the main factor seems to be legitimately low real rates, which are somewhat globally determined based on savings and investment, and with the world where it is, that's not terribly surprising.
On the VC/weighting question, that misses follow-on rounds, which anecdotally seem to drive the really extraordinary returns, in dollar but not percentage terms. You're right that the S&P skewness is somewhat tautological, but it's a very good exercise for someone who is running a concentrated portfolio of companies that aren't a big piece of the index. They have to be very confident that they're making good picks.
>If I had to guess, I would bet the Fed tries to stay neutral by buying TIPS and normal notes proportionally to the amounts they are issued, and so it's not really a factor.
But they've bought more than 100% of net TIPS issuance since the pandemic hit! It's weird.
I guess the thought experiment with TIPS is, if the Treasury doubled or halved the supply of TIPS (relative to the overall supply of debt), would that really change the market breakeven inflation rate that much? And if the answer is yes, then why even bother issuing TIPS in the first place?
I'm not sure what the original goal is, but if it's to measure inflation expectations, then it's pointless if the price is that sensitive to supply (like if breakeven inflation is 3% if supply is doubled and 1% if it's halved), and if it's to lower the govt's interest expense, then if price is sensitive to supply, they should reduce supply to drive up the price until breakeven inflation is way too high, either by the Fed buying more or Treasury issuing less.