Tech Stock Singularity

The huge SpaceX IPO has spawned the following speculation: Index funds are forced to buy shares of new large companies, in proportion to their market capitalization. That’s a classic inelastic demand. So, as the price goes up, they buy more, sending the price up even further. Maybe to infinity. The singularity has arrived.

Even the Clark Center Finance Experts Survey, in which I participate, thought this interesting and plausible enough to ask:

SpaceX IPO
a) The large demand of passive investors for shares in SpaceX in the days after the IPO will cause substantial overvaluation of the stock.

b) Rebalancing of investors' portfolios to make room for SpaceX will cause measurable price pressure on other large growth or technology stocks in the days after the IPO.

Now, you ought to be suspicious. After all, in classic finance theory, every investor holds exactly the market portfolio, just like index funds do. But there is no “overvaluation.” That theory may be incorrect as a description of the real world, but it isn’t logically wrong. “Overvaluation” can’t be automatic. There must be a Modigliani-Miller benchmark in which there is no effect at all, and getting an effect must need some Modigliani-Miller violation. But just where? It wasn’t immediately obvious to me either, so I write this post.

Start with a simple case. SpaceX’s current investors own its stock, worth $2 trillion. The rest of us own $75 trillion of US stocks, or $56 trillion S&P500. The current owners sell, say $1 Trillion of their shares to the rest of us. But where do we get the money? And what do they do with the money? Well, most obviously, we get the money by selling some of our shares of other companies. And they use the money to buy shares of other companies. In the end, after this has all percolated through many trades, they own $1 Trillion of SpaceX and an extra $1 trillion of other stocks; we own $74 trillion of other stocks and $1 trillion of SpaceX. To question b, we “rebalance” towards SpaceX, but they “rebalance” to buy the stocks we sell. The prices might not be exactly the same, as we are all better diversified. But that’s a second-order effect, and not the sort of mechanical effect that was speculated here.

The economic lesson:

• Think equilibrium, watch the money all the way to its final resting place, and remember money is a veil, all that matters is the final allocations.

You might object that SpaceX is not just selling shares so its owners can diversify. SpaceX wants to raise money for new investment, to build more rocket ships. But the same logic goes through. Suppose SpaceX is currently only worth $1 trillion. It wants to sell an extra $1 trillion of shares for a total $2 trillion valuation, and use the funds to invest in more rocket ships. The founders don’t sell any shares at all. SpaceX initially takes in cash in return for shares, and then spends the cash on rocket ships. The rockets are the assets justifying the higher ($2 trillion, not $1 trillion) valuation.

Again, though, there is a natural equilibrium in which other companies invest less. The flow of our savings that was going to finance their investment goes to SpaceX instead. Again, there is no first-order change in valuation. Slightly higher marginal product of capital in SpaceX might raise all expected returns. Alternatively, the new money might come from additional savings, but that would need expected returns on all securities to rise. Both of these effects mean a small decline in prices.

There are second-order effects. But that’s not the sort of flow effect the question has in mind.

One can go on, but I think you see the point.

Now, a contrary story:

Suppose we are all good mean-variance CAPM investors, and we hold the market portfolio. If SpaceX offers 1,000 new shares, isn’t the price indeterminate? If the price is $1 per share, $10 per share, or $100 per share, we each hold that fraction of the new market portfolio. Doesn’t that mean that demand is totally inelastic — we’ll buy 1,000 shares at any price?

In classic theory, we end up holding the market portfolio. But we don’t set out to do that. We each are trying to maximize the mean/variance efficiency of our portfolios. We read those SpaceX revenue forecasts. If the market price is $10 per share (say), we offer to buy far more than the offered shares. If it’s (say) $1,000 per share, we say no thanks. The price settles down at the present value of dividends. Since we have the same information, we all settle on the same portfolio, which is why we all settle on the market portfolio in the end. But our behavior out of that equilibrium is not passive. The demand curve for SpaceX stocks is not vertical.

Here, you see the worry about passive investing. If everybody said “well, I’ll just end up holding the market portfolio, why bother doing any research,” then indeed the demand curve would be vertical and any price clears the market. Index funds are inelastic demanders. The market depends on someone out there doing the research and offering a sloping demand curve.

Is index investing therefore bad? Lots of commenters say so. I’m suspicious. (I am one!) It’s reasonable to think that “fundamental” investors make a bit more money than we do, in return for their services. (This is also the standard theory of finance.) If more of us become passive, the returns to that activity rise. Again, don’t just think partial equilibrium. If you become passive and inelastic, someone else will become active and more elastic. The overall elasticity of demand is not obviously less because people specialize in active or passive modes. That studies of active management find so little profit suggests that fundamental analysts are doing a pretty good job and we don’t miss them.

Stocks are like the wine shop: I go down and buy by a $20 bottle of Cabernet. If I want really good wine I look in the $40 bin. I don’t know anything about vineyards and vintages. But thanks to the efforts of the wine connoisseurs who know what they’re doing and set the sloping demand curve, the more you pay the better wine you get, on average. Thanks.

It’s not obvious that a lot of dumb active money makes the market more efficient. People like me shouldn’t be valuing stocks, or vineyards.

The economic lesson.

• Don’t just think equilibrium, think about the forces supporting the equilibrium

Supply and Demand

It is true that when stocks are added to the S&P500, and index funds buy, their prices go up a bit. AI says the recent estimates are 4 to 5%. That’s a small change in expected return. A 5% price rise is a 5% decline in return, not a 5 percentage point decline in return. So if the average return was 6%, it is instead 5.7%. That’s not nothing, but it isn’t enough to set up a permanent short position. It’s in the range of the value of liquidity seen in other markets.

But that is a MM violation, needing some friction. A lot of finance now thinks a lot in terms of such static demand (or supply?) curves. You may have been tempted to start thinking “what is the latest estimate of the S&P inclusion effect?” And then contrast it with the opposite, “what is the latest estimate of the elasticity of demand for large tech stocks?” (The usual story is, after all, that selling a lot of stock all at once pushes prices down, not up.)

But demand curves live on top of basic budget constraints, MM theorems, and efficient markets. The practitioners of supply and demand finance do not think in such simple terms, with money piling up here and there not finding its eventual home, and budget constraints and equilibrium conditions forgotten. They understand “demand systems” live on top of these fundamental effects. But some readers may think all you need is a “demand curve” for SpaceX stock, untethered from substitutes and incomes, and an “inclusion effect.”

Corporate finance long hewed to a good methodology: State the MM violation. Sure, the world does not obey MM, but first understand the usually counterintuitive MM theorem (for example, firm value is the same whether the firm issues debt or equity), then understand why it might be false, and interpret facts with that eye. Often the MM theorem works better than practitioners thought. It’s a good discipline.

In any case, I bet on a price less than infinity.