The Financial Equivalent of Bungee Jumping

One quick note on my book reviews.  I have two books read, and ready to review.  I am reading Harry Markopolos’ book on Madoff, and am almost done.  That book really needed a stronger editor.  Next in the queue after that is Tony Boeckh’s new book, The Great Reflation.

I had other things planned to write this evening, but the brief plunge in the equity markets this afternoon caught my attention.  What caused the mini-panic?  Was it:

  • Dynamic hedging, a la 1987?
  • A “fat finger” placing a sell trade that was too large?
  • Panic of a single trader? Or, maybe a few?
  • Or, program trading gone awry?
  • Or, some combination thereof?

The “fat finger” hypothesis seems to be ruled out.  The NYSE says that there were no erroneous trades on their exchange, though they blame NASDAQ, and the NASDAQ is canceling trades where the rise or fall was over 60%.

My guess is that electronic trading got out of control, because human beings would not offer to sell the stock of valuable companies for exceptionally low prices, in some case less than a buck or less than a penny.

Yes, there might have been some dynamic hedging.  And some traders likely panicked and sold when they should not have.  But, buyers came in and prevented the market from tripping circuit breakers that would have shut down the markets for half an hour.  We came with a few points on the Dow of doing that.  No telling what might have happened if that had occurred.  There might have been a greater panic, or, a greater resurgence in the last half hour.  Curious that we got so close to that uncertain trigger, but did not cross the line.  More grist for the mills of conspiracy theorists.  All I can say is that I snagged some shares of Noble Corp @ $35.40, near the low of the day.

I said to one of my sons — there are four ways to act on a day like today:

  • The brave man: “Buy, buy, BUY!!!”
  • The wise man: “I will buy a little more of this undervalued stock.”
  • The indexer, or buy-and-holder: “Huh?”
  • The wimp: “Get me out of these stocks, they are killing me!”

We do not know what tomorrow will bring.  I had a very good relative value day, while losing quite a bit in absolute terms.

But my sense of the day is that some algorithmic trading programs went wild, and made trades that no sane human would.  John Henry may yet prevail in the markets.

But one note before I close: NASDAQ should not have canceled the trades.  It ruins the incentives of market actors during a panic.  Set your programs so that they don’t so stupid things.  Don’t give them the idea that if they do something really stupid, there will be a do-over.  In the absence of fraud, trades should not be canceled.

Full disclosure: long NE


  • the cancellation of the trades is a real issue. i don’t think they do it to bail out the “programs” that did stupid things – i think they do it to bail out the Mom and Pop Retail investors who don’t know how to use stop orders and ended up getting stopped out at a penny on their positions… not that that makes it any better…

  • ps – this point: “It ruins the incentives of market actors during a panic.” is simple, clear, concise, and brilliant.

  • matt says:

    There is another really good reason not to cancel the trades – integrity of capital markets. There is a large (and growing) sense that the capital markets are rigged and, given that algo traders (all large institutional investors) were the likely losers here, canceling the trades vindicates the perception that capital markets are a “Heads I win; Tails you lose” casino for the primary dealers and other large institutional investors.

    It is in the best interest financial professionals to promote the integrity of capital markets. If they don’t, it is likely that people at the bottom of the food chain (who are scalped by fees to support the securities industry) will leave these markets for good.

    Do you see gold? I do. If these crooks keep this up, the government is going to have to ban gold ownership again.

  • matt says:

    PS: I totally missed out on all of the action. I took a day trip and came home to see all of the action in hind-sight. There was a fantastic opportunity to load up on some high quality mega-cap consumer staples and utilities. THE ONE DAY I leave me terminal, I miss that.

  • wpw says:

    Usually I have some cash sitting around waiting for a chance but, because I have some things in transition, yesterday I had nothing. I was sitting here looking at my screen knowing I could buy things for 20% less than I was willing to pay, but I had nothing to buy with.

    So, go ahead cancel all those trades. We can’t have people losing money. That would be like capitalism and we cant have that.

  • Anal_yst says:

    I’m curious why the normal “Clearly Erroneous” limits don’t apply here. For most stocks in question either on the NYSE or NASDAQ that’d be 5 or 10% from the last ‘normal’ consolidated print, not 30%, not 60%. If you’re institutional and you get burned on movement like that its your own damn fault no matter how you slice it. If you’re retail and you don’t adjust your stops as the market rises (or falls), imho I have little sympathy for you (or your “broker”, but that’s another conversation). As you’ve pointed out (and the other comments echo), its called a market for a reason; actions like this do not a market make.

  • kosmik says:

    given that algo traders (all large institutional investors) were the likely losers here, canceling the trades vindicates the perception that capital markets are a “Heads I win; Tails you lose” casino for the primary dealers and other large institutional investors.

    This is actually not true. Algos are employed by tons of firms general public never heard about.

  • IF says:

    I absolutely agree with you and KD on the incentives. A mess.

  • matt says:

    “This is actually not true. Algos are employed by tons of firms general public never heard about.”

    Care to give examples of non-institutional investors who use algo trading? Because I would love to hear about small investors that are trading large enough blocks to need algos.

    Maybe my use of large threw you off (i.e., maybe you were thinking about some smaller mutual funds)? If that’s the case, it still doesn’t help the public perception that insiders get preferential treatment.

  • kosmik says:

    I can’t name our clients (we’re selling algo trading infrastructure), but typically it’s companies around 10 – 100 people which do trading for themselves or execute client orders. I don’t have any experience with hedge funds (I have a feeling they are not using algos or rely on execution services from their prime brokers). They are really a number of such companies in case of US there are a lot of them in Chicago area.

    It’s not something super-expensive which only mega-banks can afford and even high-end solutions rarely go above 1M USD per year (and it’s _high_end_ stuff for market-makers trading 24 hours across the globe).

    Also I would like point to the fact that when people talk about “algos” they don’t consider how even “manual” trading is done today. In case of fragmented or hidden liquidity and different requrements for cost/speed tradeoffs in a lot of cases it’s algo doing actual execution when people want to buy X stock. All “smart order” stuff is what broker did over phone twenty years ago, but now it’s much cheaper thanks to competiion, reduced fees spreads.

  • kosmik says:

    matt, I really don’t understand the point in bringing “small investors” into this. If you look at people who use technical analysys stuff provided by their brokers, all these leveraged FX & CFD betting etc – they are really small and they do use (simple) algos. But I don’t see anything “unfair” if such people fall behind and don’t have a change to participate on par with professional participants. I don’t see them providing financial service like liquidity providers or arbitrageurs do, they don’t have an insight into future price movements, so they don’t bring price closer to fundamentals. What they do is just noise or momentum trading.

  • Actually, we did not come close to tripping circuit breakers since they are not triggered by a 1050 point decline if it occurs after 2:30.

    A 2150 point decline after 2:00 PM leads to a market close for the day. They update the size each quarter.

    Fortunately, we do not need to think about this very often!

    Good points by all on trade cancellations and stops.

  • kosmik says:

    If you want a profile of company which might use algo, then it’s basically any trading firm.

    It may range from people who have their own view on the market and want something to automate their typical strategies, for example, pair trading or some styles of click trading.
    Then there could be all kinds of people earning from arbitrage between ETFs/futures/stocks/ADRs/options whatever.

    It could be companies doing orders on behalf of clients. They might employ execition algos allowing whem to just enter client order and forget about it without having someome to actively manage it. Such brokerage companies might sometimes go arbitrage-like strategies when they provide liquidity to certaim markets (options markets, MTFs in Europe for stocks) to get a better fee so they can do own and client orders more cheaply.

    When people talk about algos it’s all about speed. However in a lot of cases it’s just hype. Almost everyone in our domain talks about “ultra-low-latency” solutions where “ultra-low” doesn’t have any exact number behind it. But apparently not all companies using algos equally care about speed. Some favor flexibility of the systems to faster try their trading ideas, some favor throughput (example is one of our clients of much less than 70 people including backoffice who execute about 150K individual trades per day, don’t know their share on US market). Algos it’s really about automatization of work. It’s like previously cars were hand-made and now there are robots doing this. Same is in trading.

  • kosmik says:

    When I wrote about cost I meant license cost. Of course there is hardware/communications/IT cost. In case of in-house development, there is no license cost but for the development cost sky is the limit.

  • Joe Investor says:

    By attempting to eliminate risks at the individual level, nations have created macro systemic risks. We are of course seeing this play out in Europe. Canceling trades or enforcing “circuit breakers” feels very similar. By attempting to protect individual investors, who knows what the larger effects on the markets as a whole will be? Perhaps by removing natural market cleansing mechanisms (flushing out the machine trading on margin) this will be one more incremental factor in enabling the next bubble.