A SINGLE Citigroup trader caused havoc in European stocks in 2022 after the simplest of blunders: The staffer entered the value of stocks they wanted to buy into the wrong field of the bank’s trading system. The mistake triggered a five-minute selloff in European stocks from France to Norway, with US$1.4 billion worth of equities sold in error.
Citing a litany of control failures, UK regulators handed down a £62 million (S$106.6 million) fine on May 22 that, together with the trading losses from that day in 2022, mean the erroneous trade cost the desk nearly two years of revenue.
But despite the punishments (and embarrassment) that accompanies such errors, the history of modern finance is pockmarked with such mishaps as the volume of trades grows ever larger.
Earlier this year, another “fat finger” incident — trades made by human error, or even by algorithm — was faulted for tanking ASML Holding NV shares.
In 2021, traders blamed an erroneous transaction for sending Barclays Plc’s stock down almost 10 per cent, briefly trimming about £3.2 billion from the bank’s market capitalisation. And in 2019, an 83 per cent plunge and subsequent rebound in one of Singapore’s largest stocks, Jardine Matheson Holdings, left dazed traders wondering whether a “fat finger” had struck again.
The episodes serve as a reminder that even the best-designed systems can behave unexpectedly once humans enter the equation.
GET BT IN YOUR INBOX DAILY
Start and end each day with the latest news stories and analyses delivered straight to your inbox.
1. What’s a fat finger?
For the origin of the phrase, imagine someone entering a trade and pressing the button next to the one he or she meant to press on the keyboard — literally because his or her finger was too big — and you get the idea. It’s become a colloquialism that refers generally to human error, either a literal typing error or a more figurative mistake. As many financial institutions have discovered over the years, an extra zero here or there can make a huge difference.
2. Can mistakes be reversed later?
In many cases, yes. Deutsche Bank AG accidentally transferred 28 billion euros (S$43.2 billion) — far more than was due — to an outside account as part of its daily derivatives dealings, Bloomberg News first reported in April 2018. The money landed at Deutsche Boerse AG’s Eurex clearinghouse but was quickly returned.
Volume in Walt Disney shares appeared to surge for a moment in February 2015 after more than 131 million shares seemed to trade at once on the New York Stock Exchange — a transaction so big that only one shareholder probably could have placed it. But it turned out it never happened, as the initial trade order was quickly cut to a much less magical 1.3166 million shares.
3. So fat fingers don’t hurt everyone?
Not necessarily. Someone at Samsung Securities, one of South Korea’s largest brokerages, was trying to pay employees 1,000 won (S$1.20) per share in dividends under a company compensation plan in April 2018 but gave them 1,000 shares instead. That totaled about US$105 billion, or more than 30 times the company’s market value. Things got worse when 16 employees sold the stock, setting off a rout in the share price.
In December 2015, Mizuho Securities in Japan mistakenly offered to sell 610,000 shares of employment agency J-Com Co. for 1 yen each (S$0.01), instead of selling one share for the market price of 610,000 yen (S$7,460), a mistake the firm blamed on a typing error. Then problems with the Tokyo Stock Exchange’s computer system prevented the brokerage from canceling the sell order — which cost it US$345 million. Japan’s Financial Services Agency subsequently ordered Mizuho to improve its compliance and systems, as well as improve staff training.
4. Why can’t computers catch these errors?
Even the best-laid plans and systems can’t account for human fallibility. Take cybersecurity. Germany’s chief financial regulator once noted that the vast majority of computer security problems at banks are caused by “internal mishaps or mistakes or procedural problems,” rather than external forces such as criminal hackers. This has real-world consequences as regulators order banks with poor computer systems to hold more capital.
5. So how often do fat fingers strike?
Considering there are hundreds of markets around the world processing trillions of transactions at lightning speeds every day, not that often — which is why the impact is so noticeable and embarrassing when mistakes occur, as these examples show:
Gold’s no haven
Gold traders were rattled in June 2017 by a huge spike in volume in New York futures after trading jumped to 1.8 million ounces of gold in just a minute, an amount bigger than the gold reserves of Finland. Gold futures fell by as much as 1.6 per cent. One possible explanation: A mistaken trade of 18,149 lots of a futures contract, about 100 times the size of a typical trade of 18,149 ounces.
The US$617 billion bidder
In October 2014, someone placed more than 40 erroneous orders totaling 67.8 trillion yen on Japan’s over-the-counter market — greater than the size of Sweden’s economy at the time. The attempted transactions included a bid to buy 57 per cent of Toyota Motor’s outstanding shares and other big stakes in Japanese blue chips including Honda Motor, Canon, Sony and Nomura Holdings. The orders were cancelled before they were executed.
Goldman’s options disruption
Confusion swept across trading desks in August 2013 as prices for equity derivatives swung without reason, with some contracts trading for US$26 one minute and US$1 the next. The culprit was Goldman Sachs Group, which experienced software errors causing it to spew unintentional orders from the first moments of trading, Bloomberg News reported at the time.
UBS & Capcom
UBS Group AG said its Japanese unit mistakenly ordered 3 trillion yen of Capcom convertible bonds in February 2009, equal to almost US$31 billion at the time. That was 100,000 times more than it intended, according to the bank, which cited an internal system error. The trade on the Tokyo Stock Exchange’s ToSTNeT system was canceled at no cost to UBS. BLOOMBERG