Views on the integrity of global capital markets
09 September 2015

Who’s to Blame When It Rains? A Twitter-Inspired Fable about High-Frequency Trading

The following paraphrases a short Twitter discussion I had with @RemcoLenterman, @HaimBodek, and @AsselbergsSimon about high-frequency trading (HFT). It came about as a result of seeing the willy-nilly way the term was being used on the evening of ‘Black Monday’ 2015 to explain the market’s behaviour earlier that day.

Let’s say it starts raining. Clearly, umbrella sales will go up, and umbrella stores will profit. Clearly, one cannot blame umbrella stores for causing the rain.

If the umbrella store raises prices in response to higher demand, this makes sense because umbrella provision is not a public service that is to be provided at some constant price (say, like health care). However, if the umbrella store has an umbrella for sale at £10, and by the time you get to the checkout the price has gone up to £15 … this would be annoying (although perhaps less so these days in a world of Uber “surge pricing”).

If the umbrella store flips the ‘Store Closed’ sign just as you are about to enter the store that is annoying, too. However, the store has no public service obligation, and if it closes when it is raining it is not, strictly speaking, the store’s fault that you get wet.

Let’s say the umbrella store advertises it has 20 umbrellas spread over two city locations, but when someone buys two umbrellas at Moorgate, that store phones the St. Paul’s outlet, which quickly hides its umbrellas until it increases the cost on the price tags. This doesn’t seem very ‘honourable,’ but again ultimately it is the market reacting to demand. From the umbrella store’s point of view, the store doesn’t want to lose out to people walking along the street who can feel the first drops of rain and can quickly rush to buy all the umbrellas, thereby causing the store to lose out on profiting from the soon-to-materialise high demand.

Let’s say there is one particularly clever umbrella franchise that is able to predict rain very accurately and adjusts prices and umbrella supply to suit. It is never caught by surprise, and because of this it is confident enough to always offer the best umbrella prices on sunny days. Its success makes it impossible for anyone else to survive selling umbrellas in the city, and its rivals either close shop or invest in their own predictive abilities. Now, when the first drops of rain are felt by ‘scouts’ stationed at key places in the city, all the reformed umbrella stores start moving some umbrellas to storage to ensure they are not later caught short of supply, while the remaining umbrellas are quickly re-priced.

At some points in time it may even be possible that all the umbrella stores in the city are closed while they re-price their stock when it is raining. At this stage, many pedestrians may start to get nostalgic for the days when news agents and the chemist also sold umbrellas. The reformed umbrella stores would remind them that in those days the news agents’ and chemists’ umbrellas were expensive, of poor quality, and that they never had more than four or five umbrellas in stock at any time, and would quickly run out when it rained anyway.

If it starts raining, of course umbrellas are going to be in high demand, relatively low supply, and relatively high in price. This is how the market operates.

At CFA Institute, we do not think it is the umbrella stores’ fault it is raining, and it’s not a bad thing that they profit from rain. However, we:

  • Want to avoid a disorderly crush of people waiting outside an umbrella store, trying to get in, when it starts raining.
  • Are concerned when massive umbrella price swings occur, unjustified by the intensity of the rain, that start to impact other stores on the block as people have less money to spend at say the coffee shop since they ‘lost’ so much buying an umbrella.
  • Worry that a general distrust of umbrella stores will lead to a dysfunctional umbrella market even on sunny days.

The minutiae of how an umbrella store sets its prices and determines the quantity of umbrellas for sale is of little interest to anyone. But it is in everyone’s interests that, on a broad level, when it is raining, umbrellas are available.

It is difficult to determine whether this is the case when on the one hand you have drenched pedestrians complaining loudly that they are wet, and in reply, the umbrella stores saying that it is not their fault that it rained.


If you liked this post, consider subscribing to Market Integrity Insights.


Image Credit: iStockphoto.com: claudio.arnese

About the Author(s)
Sviatoslav Rosov, PhD, CFA

Sviatoslav Rosov, PhD, CFA, is an analyst in the capital markets policy group at CFA Institute. He is responsible for developing research projects, policy papers, articles, and regulatory consultations that advance CFA Institute policy positions, focusing on market structure and wider financial market integrity issues.

8 thoughts on “Who’s to Blame When It Rains? A Twitter-Inspired Fable about High-Frequency Trading”

  1. nanex says:

    Analogy misses:
    1) Umbrella manufacturing is publicly subsidized
    2) Authorized umbrella sales only
    3) One can induce rain (it costs a lot, but it works)

    1. Sviatoslav Rosov, PhD, CFA says:

      Thanks for your comments, have enjoyed reading your blogs as well. Agree number 3 is very undesirable. Longer term I think a decision will have to be made about number 1. Is this actually a public utility?

  2. David Gordon says:

    Excellent post! Almost sounds like umbrellas should become a public service…

    1. Sviatoslav Rosov, PhD, CFA says:

      I think at some point regulators will have to make a decision along what welfare dimensions they want the market to be prioritized and go with that. Some will lose out.

  3. Emerich says:

    This fable is problematic to the point of incoherence.

    1. The whole point of peak pricing is that you don’t get “a disorderly crush of people” waiting outside to buy umbrellas. The peak prices will induce some people to wait a while at home or in their offices while it’s raining; or stop at a coffee shop to wait out the downpour if they can spare the time; or perhaps, carry their own umbrellas more regularly to avoid paying peak prices. Such behavioral adjustments will in themselves ameliorate the peak prices. For those people without umbrellas who must get from A to B during the downpour, the peak prices will ensure that umbrellas will be available and without long queues. Again: the congestion pricing ensure that supplies are available, typically without queuing.

    3. Some umbrella shops or chemists or what have you might discover that keeping a fixed price for umbrellas in all weather builds customer trust and loyalty. But would we want to require all umbrella vendors to adopt that business model? Why on earth should we? The fact that umbrella vendors generally do not employ congestion pricing is in itself the reply to your fable. Vendors generally believe the advantages of such a pricing model are outweighed by the disadvantages, perhaps including reputational loss. So what’s your worry?

    4. It’s absurd that “At some points in time it may even be possible that all the umbrella stores in the city are closed while they re-price their stock when it is raining”. The whole point of repricing for peak demand is to sell product during that peak demand. You’re assuming all the shopkeepers are overcome by gross incompetence and/or stupidity at the same time. The fact that that doesn’t happen is evidence that shopkeepers are generally sensible. Those who aren’t don’t stay in business long.

    I know this is a “fable” that’s supposed to be about HFT. But this suggests perhaps you don’t understand HFT any better than you do price theory.

    1. Sviatoslav Rosov, PhD, CFA says:

      Agree fable isn’t perfect!

      Regarding your point 1, I agree and I think that is what the first few paragraphs of the blog post highlight. Regarding your point 3, the issue I was raising was there is some concern that liquidity provision has become homogenised, nothing to do with fixed prices or price theory. Regarding your point 4, I have been shown on the “Black Monday” tape Costco trading with an 8-cent spread because there were no bids so I think the metaphor holds up.

      I get the impression you think the piece is anti-HFT, but it certainly isn’t. Like the blog post says, we are interested in the market working well. The Dow going down by 1,000 points then going back up 600 points doesn’t “seem” to fit this description.

  4. Stefan says:

    I’m very sorry, but there is a very big difference between how HFT algos operate and how “real economy” service provider operates. Additionally, you gloss over some issues that are very explicitly described in the ethical sections of the CFA section, from the level 1 exam onward.

    The appropriate analogy is that HFT algos are umbrella market observers, that see a customer intending to buy the full stock of umbrellas from several places on the same day (regardless of whether it rains or not, there could be various underlying economic motives).

    The market observer then beats the buyer to the other sellers (he has bought express highways between venues), buys their stock and sells it on to the customer at a slightly inflated price (it can’t set the price arbitrarily high, it might just mean the customer cancels the order).

    There are 3 issues with this practice why it is unlikely to occur in real life:
    1) Price collusion between independent companies is illegal
    2) Monopolies are illegal
    3) Its prohibitively expensive to buy the full stock of umbrellas in sufficient quantities and simultaneously to set the price at a sufficiently high level to make a profit, thus rendering the trade uneconomic.

    The first 2 conditions are necessary to control the price and quantity in the entire market place but are actually illegal. The only reason why it is possible for HFT algos to make a profit is because:

    1) there is a principal/agent relation between ultimate capital owners and their brokers that trade on their behalf, potentially creating a conflict of interest in relation to the price paid for securities (would the client mind paying an additional 0.01% in transaction cost as they intend to hold the position for years?).
    2) due to the electronic nature of trading nowadays, it is possible to achieve the necessary scale to make a profit from a marginal increase in price.

    In my view you should provide a slightly more nuanced picture for analogy, because the current version appears very pro-HFT..

    1. Sviatoslav Rosov, PhD, CFA says:

      Agree the analogy isn’t perfect, that was not really the point. Much better explanations of HFT are available elsewhere that is true.

      The blog post was really written to highlight the fact that there is clearly confusion about what the market should be doing, who it should be serving, and how this is to be achieved. This is not helped by each side of the debate having clear, vested interests.

      Interestingly you say it comes across as pro-HFT while the pro-HFT guys have told me it comes across as anti-HFT!

Leave a Reply

Your email address will not be published. Required fields are marked *



By continuing to use the site, you agree to the use of cookies. more information

The cookie settings on this website are set to "allow cookies" to give you the best browsing experience possible. If you continue to use this website without changing your cookie settings or you click "Accept" below then you are consenting to this.

Close