How to lose money - fast!

Summary

Podcast Summary: More or Less – "How to Lose Money – Fast!"

Date: August 13, 2012
Host: Tim Harford (BBC Radio 4)

---

Episode Overview

This episode of "More or Less" explores the world of high frequency trading (HFT) and how an ordinary software error at a trading firm led to the catastrophic loss of nearly half a billion dollars in just 45 minutes. Tim Harford, together with financial experts Felix Salmon, Donald McKenzie, and Andrew Haldane, investigates the mechanics, risks, and potential safeguards around super-fast, computer-driven trading on the stock market.

---

Key Discussion Points and Insights

1. The Knight Capital Disaster (00:13–01:32)

• Tim Harford recounts the recent Knight Capital incident, where a faulty HFT algorithm led to catastrophic financial loss.
  • Felix Salmon's account:
    "They switched it on and immediately they started losing literally $10 million a minute... This went on for 45 minutes. And at the end of it all, they wound up having lost $440 million." – Felix Salmon [00:48]
  • The error was due to the algorithm repeatedly buying high and selling low, executing thousands of losing trades per minute.

2. How High Frequency Trading Works (01:32–03:12)

• Automation and Speed:
  HFT relies on computers to execute trades at speeds and volumes impossible for humans.
  • Harford illustrates HFT speed by comparing it to slowing down time in the 100m sprint, showing just how quickly trades occur relative to human reaction.
• Efficiency vs. Risk:
  High frequency trading is generally more efficient than human trading, but errors or excess speed can have disastrous consequences.

3. The Flash Crash of 2010 (03:56–05:15)

• On May 6, 2010, the Dow Jones dropped 600 points in five minutes and recovered ten minutes later with no obvious cause.
  • Felix Salmon:
    "There was this huge crash in the stock market for no reason. And then 10 minutes after that, it went back up again and no one knew what had happened." [03:56]
• A large, poorly executed trade triggered a frenzied response from algorithms, leading to a rapid market collapse and bounce-back.
• Human operators pulled the plug, resulting in a sudden liquidity vacuum.

4. Circuit Breakers and Human Intervention (05:15–05:57)

• The crash ended when the Globex exchange’s "StopLogic" feature automatically halted trading for five seconds.
  • Donald McKenzie:
    "That was in fact enough time to halt the downward spiral. It gave human traders... the chance to just take a quick glance at the rolling news feeds... and then to think, well, perhaps this plunge of prices is a buying opportunity." [05:22]

5. Anatomy of Trading Algorithms (05:57–07:19)

• McKenzie outlines types of trading algorithms:
  • Protective algorithms: Prevent loss when executing human trades.
  • Market-making algorithms: Profit from small price spreads.
  • Statistical arbitrage strategies: Exploit short-lived discrepancies in asset relationships.
• Some algorithms are "predatory," such as "algo sniffers" that anticipate and exploit slower trading programs.
  • The market has evolved into an "arms race" of ever-faster, more complex algorithms.

6. The Evolving Nature and Risks of HFT (07:19–08:09)

• Andrew Haldane (Bank of England):
  • Describes the HFT landscape as a "complex array of multiple mutating, interacting machines" with algorithms that can become obsolete within weeks.
  • Warns:
    "That's not an accident waiting to happen, that's an accident that has been happening with increasing frequency over the last few years. We shouldn't wait for the equivalent of the space shuttle disaster before remedying the situation." [07:39]
• Haldane points out that warning signs are already visible and urge action before a major catastrophe.

7. Potential Solutions and Safeguards (08:09–08:46)

• Tim Harford: Asks Haldane for action steps.
• Andrew Haldane:
  • Suggests simple mechanisms, like circuit breakers, to "stop the clock" temporarily when markets spiral, allowing for reassessment and manual intervention.
    • "Circuit breakers when spiraling is happening can help prevent a bad situation becoming calamitous." [08:33]

---

Notable Quotes & Moments

• Felix Salmon:
  "They switched it on and immediately they started losing literally $10 million a minute." [00:48]
• Tim Harford (on HFT speed):
  "Usain Bolt is understanding that it's time to run an algorithm that, written into a high frequency trading platform, could complete about 165,000 separate trades." [02:29]
• Andrew Haldane:
  "This is a system that's constantly evolving, that's traveling at ever higher velocities, and it's just difficult to know what will pop out next." [07:27]
  "We already have enough lights on the dashboard flashing red to want to do something differently." [08:01]
• Donald McKenzie (on the circuit breaker):
  "That was in fact enough time to halt the downward spiral." [05:27]

---

Important Segment Timestamps

• Knight Capital disaster explained: 00:13–01:32
• What is high frequency trading?: 01:32–03:12
• The 2010 flash crash: 03:56–05:15
• Circuit breakers and market safeguards: 05:15–05:57, 08:09–08:46
• Anatomy of trading algorithms: 05:57–07:19
• Risks and future of HFT: 07:19–08:46

---

Tone and Language

The episode balances technical explanation with vivid analogies and a light, wry tone, characteristic of Tim Harford’s style. The contributors speak frankly and plainly about the risks, using analogies and plain English to demystify the jargon of high-frequency finance.

---

Conclusion

This episode offers a sharp, well-paced primer on the perils and peculiarities of high frequency trading, combining memorable storytelling (e.g., the Knight Capital disaster), accessible metaphors (Usain Bolt race), and sober analysis of future risks and remedies. For those new to modern trading markets, it’s an essential, engaging listen.

Transcript

[BBC Announcer]

Thank you for downloading from the BBC. For details of our complete range of podcasts and our terms of use, go to bbcworldservice.com podcasts

[Tim Harford]

hello, and welcome to More or Less on the BBC World Service. If there was an Olympics for number crunching, we'd be striving for gold. Let me tell you a story about a strange thing that happened last week. The New York Stock Exchange launched a new electronic trading platform. In preparation for this, a company called Knight Capital had created some software to link up with a new platform in order to trade shares on it. The stock market opened and Knight Capital prepared to launch its new software.

[Felix Salmon]

There was some problem with the program. We don't know exactly what this is.

[Tim Harford]

Felix Salmon, finance blogger for Reuters, based in New York.

[Felix Salmon]

They switched it on and immediately they started losing literally $10 million a minute. It looks like what they were doing was they were buying high and selling low many, many times per second and losing $10 or $15 each time. And this went on for 45 minutes. And at the end of it all, they wound up having lost $440 million.

[Tim Harford]

Oops. But how could a company lose so much money so quickly? This is the latest chapter in the story of something called high frequency trading. Investors have always valued being the

[Unknown Caller]

345-612-3345.

[Tim Harford]

But high frequency trading is something different. Automatic trades conducted by computers, each one racing to be first. Humans still watch the systems, but the computers move far too quickly for us to react to everything they do. And at night, Capital. The computer glitch meant the company was making trades it didn't intend to make. That's how to lose almost half a billion dollars in less than an hour. To give you a sense of how fast high frequency trading can be, imagine slowing down time so that 10 seconds lasts 30 minutes. In this parallel universe, Usain Bolt ran the hundred meters in just under 29 minutes. It took him 30 seconds just to react to the starter's pistol. And the pistol bang itself lasted two minutes. The lets play you just the B of the bang. Just the beginning of a two minute gunshot sound. And now let me play you some pulses. One pulse for each trade from a high frequency trading computer. Now that doesn't work. There are still too many trades in the 1600ths of a second. Usain Bolt is understanding that it's time to run an algorithm that, written into a high frequency trading platform, could complete about 165,000 separate trades. That's pretty fast. So let's slow it down even more. The 10 second hundred meter race now lasts two months. The 160 millisecond reaction time lasts just over a day and the bang of the starter's gun lasts four and a half days. On that incredibly slowed down timescale, here's how quickly the high frequency trades are happening. And it's getting faster all the time. Now, this isn't quite as insane as it sounds. These computers, all competing with each other are a lot cheaper and more efficient than human traders trying to match bids to buy and offers to sell. So, within reason, automated high frequency trading is a good thing. But it's possible to have too much of a good thing. On 6 May 2010, the UK was preoccupied with a general election. But on the other side of the Atlantic, a very different story was unfolding.

[Felix Salmon]

We were all sitting around in the middle of the afternoon on a relatively slow news day, and suddenly the Dow Jones Industrial average was down 600 points in a matter of five minutes. There was this huge crash in the stock market for no reason. And then 10 minutes after that, it went back up again and no one knew what had happened.

[Tim Harford]

What may have happened was, was that somebody rather clumsily tried to make a big trade all of a sudden on an electronic exchange called Globex. As the price dropped sharply in the process of trying to find willing buyers, the algorithms of the high frequency traders plugged into Globex went into overdrive. And the flash crash wasn't just what happened when the algorithms were in a frenzy. It was also what happened when human beings pulled the plug on all of those individual algorithms.

[Felix Salmon]

A whole bunch of high frequency traders saw a lot of weird stuff going on in the market, which didn't make any sense. And they said, you know what, we don't know how to make money in this. This is outside what our algorithms are used to. We're all just going to hit our off buttons at the same time. And what that did was it meant there was no what's known as liquidity in the market. There was no one in the market buying and selling, because the market has become reliant on these high frequency traders to buy and sell stocks whenever anybody wants to trade.

[Tim Harford]

What actually stopped the flash crash was absurdly simple. The exchange itself, Globex shut down.

[Donald McKenzie]

How it stopped was that Globex has got a program built into it called the StopLogic functionality. And in this case it stopped the system trading for just five seconds. But that was in fact enough time to halt the downward spiral. It gave human traders, for example, the chance to just take a quick glance at the rolling news feeds discover that nothing catastrophic in the wider world seems to have happened, and then to think, well, here perhaps this plunge of prices is a buying opportunity.

[Tim Harford]

That's Donald McKenzie, a sociologist at the University of Edinburgh who studies financial markets. Professor McKenzie has been interviewing dozens and dozens of people in the algorithmic trading industry, and he divides the flash traders into, into categories. First, there are algorithms designed not to lose money while executing a trade that's being placed by a human. Then there are algorithms designed simply to make money by finding buyers and sellers with a small margin between them. Third, there are algorithms which find statistical relationships between different shares or bonds. And when the statistical relationship fails to hold even for a moment, they jump in and make a bet that normal service will soon be resumed. Now, there's nothing immoral about these three approaches, but there are more predatory strategies. One is called algo sniffing. Here a super fast computer tries to find other computers going about their everyday business of buying or selling shares and figures out what they're going to do and when. The algo sniffer can then get ahead of the game and exploit the slower computer. And of course, you can have algo sniffer sniffers and algo sniffer sniffer sniffers. In a high frequency arms race. No wonder speed can be so important. Andrew Haldane, executive director for Financial stability at the bank of England, is increasingly interested in how high frequency trading works and what the future might hold.

[Andrew Haldane]

What we have out there now is this complex array of multiple mutating, interacting machines, algorithms. The half life of an algorithm might be as little as a few weeks. So this is a system that's constantly evolving, that's traveling at ever higher velocities, and it's just difficult to know what will pop out next. And that's not an accident waiting to happen, that's an accident that has been happening with increasing frequency over the last few years. We shouldn't wait for the equivalent of the space shuttle disaster before remedying the situation. We already have enough lights on the dashboard flashing red to want to do something differently.

[Tim Harford]

Presumably you're in a position to have a lot of influence. You're in charge of financial stability at the bank of England. What would you recommend?

[Andrew Haldane]

There's at least two or three relatively simple things that could be done. I think one of the most straightforward would be on occasions when things are spiraling out of control, we ought to be able to stop the clock. That timeout period will allow prices to be reset, algorithms to be adjusted, perhaps even turned off, and the market to restore some semblance of stability, so circuit breakers when spiraling is happening can help prevent a bad situation becoming calamitous.

[Tim Harford]

Andy Haldane, Executive Director for Financial Stability at the bank of England and that's all we have time for this week. Please keep your questions, comments and suggestions coming in to more or lessbc.co.uk. our brand new website with print versions of our best stories [e-mail] more or less. Please check it out and perhaps subscribe to our podcast. Until next week. Goodbye.

[Program Credits Narrator]

More or Less was presented by Tim Harford, the Financial Times undercover economist. The producer was Richard Knight and the editor was Richard Varden.

[BBC Announcer]

There are dozens of different podcasts now available from the BBC, including news, documentaries, science, business, arts and sports. The details of them all go to bbcworldservice.com podcasts.