selling volatility

It occasionally happens that, for some exogenous reason, some market participant or group of participants sells a lot of options on a stock, and it is then noted that this means the sell-side of Wall Street is often involuntarily buying them, which typically means that those sell-side shops will go out and delta-hedge; since these move one-for-one with the stock when the stock price is high, and don't move when the stock price is low, that entails buying the stock as it rises (nearish the strike price), and selling as it falls, thereby stabilizing the price.  

All of this is well and good, but I'm inclined to skip some steps; the buy-side agents are selling volatility, and that should generally push down the price of volatility.  In most markets, selling a thing pushes down the price, and when ordinary supply and demand work reasonably well, reequilibration at the lower price will entail moving to a situation in which the marginal cost of the thing is lower than it was at the beginning, and the detailed mechanics don't really matter that much.  "Volatility" may seem abstract, but it works the same way.

An important special case here is the mortgage bond market; homeowners there have (and to some extent use) an option on interest rates, and when those become more likely to be exercised, that tends to destabilize long-term interest rates.  Again, you can walk through a discussion of interest-rate investors trying to match shrinking durations, and eventually you find dropping interest rates being pushed lower and rising ones pushed higher, but if you're just interested in the dynamics of interest rates themselves, you can skip all these steps; a bunch of people who have no idea what duration, convexity, or volatility mean are unwittingly long volatility, and that pushes up prices.

I'm posting this at the moment because Matt Levine is writing about greenshoes, which are options granted to underwriters of IPOs, for the purpose of enabling the underwriter to stabilize the price of the IPO.  The details of the mechanics are a bit different from what you might think based solely on that sentence, but, as I keep saying, it's not that sensitive to mechanics; if the company thrusts options into the financial system, that has the effect of pushing down volatility.  Some people don't like greenshoes, and a problem with them that Levine argues is probably hypothetical is noted in his newsletter.  It has made sense to me for a long time to let locked-up shareholders sell at some minimum price before the lockup period ends, e.g. you IPO at $40 and tell certain classes of shareholders that they can't sell below $60 in the first six months, rather than tell them they can't sell at all.[1]  I wonder now whether it makes sense to distribute warrants with the initial share allocation.[2] [3]

It's hard to remember one's previous state of ignorance, especially when it's been so long, but I believe that 20 or 25 years ago I assumed an IPO worked more or less like a direct listing with a capital raise; until much more recently I didn't realize that those weren't even allowed.  (I became aware of that when discussion of changing that heated up, and it looks like in 2021 it will finally be permitted to do what I assumed in the 1990s was the way a company went public.)  I mention this in substantial part by way of warning that the hard parts of finance are legal and customary, where, as is true of most people, by "hard parts" I mean "the parts I don't know and/or understand so well."

---

[1] In a certain kind of theory, this should make no difference; forward-looking investors on the day of the IPO know these shares will be available to the supply side of the market in six months, and should price that in.  A lot of my practical concerns with ideas I'm putting forward in this post are related to the extent to which this fails in practice, and the related fact that the amount of genuine "float" in the early days of a new issue is quite small.

[2] In this case the size of the early float of shares endorses my proposal, insofar as the reasons the float is so small are likely not to apply to these warrants.

[3] One of the complaints about the greenshoe is that companies don't like not knowing how many shares they're selling; as I've argued in some previous posts, a lot of companies should be offering more of a bona fide, upward-sloping supply curve anyway.  If the warrants execute, it's likely the investors think the company can make good use of their additional dollars, while if they don't, they don't, and the company should have some idea what it would best do with the additional investment should it be forthcoming.  It is perhaps worth noting here that if demand is volatile, an elastic supply curve is a way of absorbing some of that volatility into quantity rather than price; if you don't like talk of "selling volatility", perhaps the fact that a warrant exercise increases the quantity supplied is more intuitive.

index minus

In a certain kind of ideal market, securities lending fees should be zero; at any positive price, there would be effectively infinite supply.  In actual practice, a lot of shareholders, for institutional or other reasons, don't lend out their shares.  Those that do lend out shares tend to be price-inelastic, at least above some de minimus price; they'll lend out all of the shares they have available.  Retail borrowers typically don't directly pay a lending fee; they pay it indirectly in various ways, and those ways tend not to respond to the level of the actual fee; if the actual fee is higher than what the broker is getting from the customer, they are simply unable to borrow.  Retail demand, thus, is also very inelastic for most possible prices, but with a dramatic elastic region around a particular price that depends on things like nominal interest rates.

Index investing has grown a lot in the last generation (after growing a lot in the generation before that); a lot of people simply own the S&P 500 in some literal or practical sense.  The saver owns the same fraction of every company in the index.  Some retail investors have, for a while, adopted an "index plus" strategy of putting most of their money in a passive index fund but putting a little bit in stocks that they select.  Increasingly, people are interested in something like an "index minus" strategy; they are largely agnostic about most stocks, except that they think they should own "stocks", but there are a few stocks that they think are overpriced, or think are bad for ESG-related reasons.  My impression is that, at least at the retail level, they mostly end up just buying an ESG fund, or buying an S&P fund and perhaps selling a couple of stocks short, and that that's about the best they can do.

I'm trying to work out in my head whether there's a sensible market for an index minus brokerage-type account; the broker here would guarantee that a customer could in some sense "sell short", but only in the amount that that customer owned the stock.  To the extent that different customers "short" different stocks, part of the custodial assets of the company could be in a Vanguard S&P fund or do something with futures.  To the extent they don't, the company might have to do some of the work of a fund itself,  perhaps including some securities lending, and would have some of the same economies of scale that most mutual funds do.  I think this ends up looking a bit like a subsidy to customers selling hard-to-borrow securities, but the limit on the size of the sale keeps that manageable, and it may be an attractive enough idea to enough people looking to put their money in stocks, many of whom would only sell cheap-to-borrow stocks anyway, that there's a business opportunity to be had here anyway.