New Way to Bet on Commodities: Just Bet on Commodities

Two-way bets on natural gas, or the VIX, are a whole lot easier than trying to store natural gas, or the VIX for that matter. Where would you put the VIX?
Why not try some synthetic fracking? Photographer: Ty Wright/Bloomberg

Here is a neat thing that is also an investment proposition. I have no opinions on it as an investment proposition, but I'll pass it along as a neat thing. 1 A company called AccuShares has built a series of exchange-traded funds that let you bet on or against various commodities (plus the VIX). There are lots of such exchange-traded funds, but as far as I know they tend to follow one of two models:

  • Take investors' money, buy pile of stuff, use pile of stuff to back investors' claims, or
  • Take investors' money, buy commodity (or VIX) futures, roll those futures, try to track some rough approximation of the commodity's price.

Each approach has problems. Buying a pile of stuff requires storing and guarding that stuff, and only works for things such as gold that are non-perishable and relatively easy to store. Buying futures requires rolling them when they expire, adding trading costs and tracking error and tax inefficiency and so forth.

The AccuShares approach is:

  • Take investors' money, in equal amounts from long and short investors.
  • Do nothing.
  • If the commodity price -- based on some observable index -- goes up, give some of the short investors' money to the long investors.
  • If it goes down, give some of the long investors' money to the short investors.

Isn't that neat? My favorite part is probably the 45-75 basis point management fee -- for doing nothing! 2 -- but there are other clever bits too. For instance, one obvious problem with this strategy is that it keeps paying out money to the winners every quarter, 3 so you eventually run out of money, making the fund sort of impermanent. They have some clever strategies there, including paying distributions in paired long and short shares if they're getting close to running out of money, but realistically the strategy for these things to work involves mostly getting a lot of people to invest in them. 4

This presses a lot of my buttons. 5 For one thing, it is a logical next step in the continuing abstraction of finance: Whereas once you invested by, like, giving money to your uncle to open a general store, now you invest by saying, "What exposures to the world do I want?" and then pressing some buttons on your computer to get those exposures. Imagine, once upon a time, to invest in oil, you had to buy oil! And then put it somewhere. A messy business. This is just pure button-pushing. No oil or gas or livestock are involved; it relies solely on the expectation that some people want one abstract exposure, and others want the opposite exposure.

For another thing, there is a pleasing retailization of derivatives here. Institutional investors have long had lots of ways to just make bets. Want to bet on interest rates? Try an interest rate swap. Want to bet on a stock without owning the stock, for some reason? An equity total return swap is just the thing for you. Want to bet on a stock, like Alibaba, that doesn't actually trade yet? There's something for that too.

There is less of this for retail investors. Not none, exactly; you can sort of find yourself in a total return swap with Credit Suisse if you're not paying attention. But generally speaking, if you go to your stockbroker and say, "I would like to bet on things I don't own," your broker will suggest you try your luck at a casino. 6 Now he can suggest these things.

Relatedly, I've been a fan of bucket shops ever since I learned what they were from Jesse Livermore. The bucket shop was an old-timey place where people went to bet on stock prices, without ever owning the underlying stocks. Also, there was probably a lot of fraud, but that's not intrinsic to the model. The model is just letting people bet on financial quantities without the bother of owning them. And now it's back! AccuShares is both a step forward into the future of dematerialized finance, and a quaint throwback to the olden days before finance was tamed.

If you wanted, you could of course tell negative stories. You might say that it is good that it's difficult for retail investors to speculate on commodity prices, and that more commodity speculation, and more opportunities for retail investors to take exciting new risks, are not what our financial system need. I'll just leave that one here for you to think about.

Or you might find all this abstraction too much. AccuShares lets you invest in natural gas, say, but not by buying natural gas from companies that drill for natural gas. Your investment is never put to productive use, never reinvested in more drilling. It just goes into a pot of money to eventually be split by you and the guys betting against you. It hardly even sends a price signal: If a lot of people buy AccuShares long-natural-gas shares, that won't directly push up the price (or supply) of natural gas; it may just push up the supply of AccuShares. 7 Not everyone shares my fondness for finance that abstracts dizzily away from the physical messiness of reality. Some people prefer finance that has an effect -- possibly even a good one -- on the real world.

This column does not necessarily reflect the opinion of Bloomberg View's editorial board or Bloomberg LP, its owners and investors.

  1. Disclosure! It's a thing cooked up by friends of a friend of mine. This is so not investing advice! But it never is.

  2. Unfair; the managers have to invest the proceeds in cash equivalents, and figure out all the distribution-y stuff. I just like to think of them charging a fee for keeping their zen-like calm in the middle of frantic long and short commodity speculators.

  3. Plus sometimes when there are big enough moves between quarters. There are a lot of rules, read them here.

  4. A general cleverness question is, does this work? As like a no-arbitrage ETF? You have to read the prospectus to really get it, but the gist is:

    • There are equal numbers of Up and Down shares.
    • There's a single net asset value, and Up and Down share in it based on whether the underlying is up or down.
    • If Up shares trade above NAV, while Down shares are in line, then a market maker can create new units by delivering cash to AccuShares, getting new Up and Down shares, and selling the Down shares for fair value and the Up shares for more than fair value.
    • If Up shares trade too high, say, and Down shares trade too low, such that the thing overall trades in line with NAV, then there's no delivery arbitrage, which is what usually keeps ETFs honest. So you're left with statistical arbitrages: Either you just buy Down shares, sell Up shares and wait for the prices to converge due to the cash distributions; or you buy Down shares, hedge by buying the underlying commodity (or futures, etc.), and make market-neutral profits if and when the Down shares trade back in line with the commodity.
    That last part is sub-optimal but maybe OK.
  5. One minor one that has been a recent obsession: This thing illustrates -- I mean, assuming anyone buys it -- the value proposition of the middleman who provides liquidity. If I want to bet that oil prices will go up, and you want to bet that they'll go down, we could just make a bet, between us. That would be free. Instead we'll -- maybe! -- pay AccuShares 75 basis points a year to manage our bet for us. Why? Well, because this way we don't have to find each other. AccuShares -- and the brokers buying creation units -- takes care of that for us. For, you know, a price.

  6. Or with listed options, but come on. Leave that to the insider traders.

  7. Maybe? See note 4. This is opaque; obviously one possible arbitrage involves propagating the price signal from AccuShares to the actual commodity price.

(Matt Levine writes about Wall Street and the financial world for Bloomberg View.)

To contact the author on this story:
Matthew S Levine at

To contact the editor on this story:
Zara Kessler at

Before it's here, it's on the Bloomberg Terminal.