Market manipulation has been on my mind a lot recently so I guess it's worth explaining what it is. In simple form, market manipulation is:
- buying a lot of an asset,
- with the goal of pushing its price up,
- so you can sell it at that higher price to a price-insensitive buyer.
Or vice versa, with selling. If you think you've spotted market manipulation, a useful question to ask is, "who is the price-insensitive buyer?" Because buying a lot of an asset with the goal of pushing its price up and then, just, like, selling a lot of it on the market is not in general a positive expectation strategy.
You know who is a great price-insensitive buyer? A major central bank that has embarked on a program of quantitative easing. The goal of quantitative easing is to push bond prices up (interest rates down), so sort of by definition the bank is not trying to buy bonds cheaply. If it was trying to buy bonds cheaply, it would not announce the quantitative easing in advance. Really, its goals are the opposite: Not only would it rather push prices up than buy cheaply; it would rather push prices up than buy at all. If the Fed could say "quantitative easing, go, move rates down by 50 basis points," it would be happy to do that without buying bonds.
Here is a U.K. Financial Conduct Authority action against a former Credit Suisse trader who manipulated gilt prices during a Bank of England quantitative easing auction in 2011. The trader, Mark Stevenson, was fined 662,700 pounds and banned from the industry, and, fair enough, he did bad:
Mr Stevenson, an experienced bond trader formerly employed by Credit Suisse Securities (Europe) Limited ("CSSEL"), bought £331 million of the UKT 8.75% 2017 (the "Bond"), a UK government gilt, between 09:00 and 14:30 on 10 October 2011. The Bond was relatively illiquid and Mr Stevenson's purchases represented approximately 2,700% of the average daily volume traded for the Bond in the previous four months and 92% of the value of the Bond purchased in the IDB market on 10 October 2011. The price and yield of the Bond significantly outperformed all gilts of similar maturity on 10 October 2011, as a direct result of Mr Stevenson's trading.
He then tendered it into the quantitative easing auction at the end of that day, and ... oops! The BOE decided to be price-sensitive, for the only time ever, because this bond's price got really out of whack, and because other dealers called the BOE to complain, as you'd expect them to:
By 09:39 (38 minutes after Mr Stevenson began trading in the Bond), a market participant had telephoned the BOE regarding the Bond's outperformance. Several other market participants telephoned the BOE throughout the day, suggesting that the Bond had been "squeezed", "rammed", and that someone "was messing around with" it.
No honor among thieves etc. So the BOE didn't buy any of the bond that Stevenson was trading, its price got right back into whack, and Stevenson lost a lot of money:
Remember up is down, so that blue line going down at the beginning of the day was Stevenson pushing up the price of his favorite bond, and that line shooting up at the end of the day was him realizing that he'd made a terrible mistake and dumping his bonds until they got back in line with all the other bonds. You can't tell for certain, but it sure looks like he sold his bonds at lower prices (higher yields) than he bought them at. Also: Fined and banned from the industry!
A bigger lesson is that hogs get slaughtered. The FCA's notice is full of suggestive quotes and they are ... well, let's read them.
During the day, Mr Stevenson took part in a telephone conversation with Broker A, a friend of Mr Stevenson and an agency broker. During the telephone conversation with Broker A, Mr Stevenson said "And we've been loading up with QE trades for months", and "QE's are ... cake...". The Authority has concluded that Mr Stevenson was indicating his belief that QE was an opportunity to profit from selling gilts to the BOE.
QE was an opportunity to profit from selling gilts to the BOE! The BOE wanted market makers to go out and buy bonds and then sell them to the BOE. If you want a market maker to provide you with liquidity, you pay for that.
Mr Stevenson received a telephone call on 7 October 2011 from Risk Management at CSSEL, who were calling in respect of a profit and loss spike on a book jointly held by Mr Stevenson and another trader. Mr Stevenson explained on the telephone to the Risk Management officer that:
"... I mean the MPC, the Bank of England has announced you know they're extending their asset purchase scheme yesterday, and you know we're preparing to... well we're buying some assets to sell to them, I mean basically that's what's going on."
That basically is what was going on! Again, that was the mechanism of quantitative easing: The Bank wanted market makers to buy some assets to sell to the Bank. Because it wanted the bonds, sort of, but really because it wanted the market makers to push the prices of those assets up. This didn't quite work on October 10 -- all those bonds in the chart ended the day at lower prices (higher yields) than they started -- and is phrased euphemistically in the FCA Notice, which refers to "measures designed and implemented by the BOE to stimulate the economy." But they were meant to stimulate the economy by pushing bond prices up.
Every bank that sold bonds to the BOE that day was doing something similar to what Stevenson was doing, just in less stupid and piggish form. Buy some of a bond, not all of it. Push its price up a bit, not a ridiculous amount. And then sell to the BOE at a small profit, not a silly one.
The FCA's reasons for punishing Stevenson include:
This trading had the potential to impact upon taxpayers, with the BOE potentially overpaying for the Bond and HM Treasury indemnifying the BOE for losses on gilt holdings purchased during QE.
But since the goal of QE was to raise the price of gilts, this is a somewhat conceptually puzzling complaint. The whole idea was to overpay, relative to unaffected market prices.
Here's a better reason:
The market abuse had a significant impact on the market on 10 October 2011 and risked undermining market confidence and having a serious adverse effect on the orderliness of the market. The trading took place on the first day of the second round of QE (which was designed and implemented to stimulate the economy) and is the only time the BOE has rejected all offers in a bond. Some market participants refused to trade the Bond at all on 10 October 2011 because of the price movements and others may have suffered losses as a consequence of the decision of the BOE to reject all offers in the Bond.
That's actually two reasons. One, market manipulation is bad because it reduces confidence in markets, which is fair enough. But, two, QE manipulation is bad because it reduces confidence in QE. You don't want dealers who buy bonds to sell at a profit to the Bank of England to lose money because Stevenson bought too many bonds to sell at too big a profit to the Bank of England. If dealers can't trust that buying bonds for QE will be predictably profitable, they won't do it, and QE won't be effective in pushing up bond prices. Which is reason enough to punish Stevenson for pushing up prices too much.
This column does not necessarily reflect the opinion of Bloomberg View's editorial board or Bloomberg LP, its owners and investors.
(Matt Levine writes about Wall Street and the financial world for Bloomberg View.)
Possible answers include customers who've agreed to sell at the 4 p.m. FX fix, say, or derivatives counterparties.
Incidentally here is the BOE's price sensitivity, from paragraph 4.12 of the notice:
The BOE purchased eligible gilts in QE through a competitive reverse auction process. Between 14:15 and 14:45 on a QE reverse auction day, GEMMs were invited to submit offers to sell gilts to the BOE. This was a competitive process, with GEMMs stipulating a price they wished to achieve for the sale of the gilts. At 14:45, the 30 minute reverse auction window closed and the BOE took a snapshot of the DMO mid-market price for each gilt eligible for purchase on that day. All offers received by the BOE from GEMMs were then value-ranked relative to the mid-market price for the relevant gilt, with the cheapest offers being purchased first and more expensive offers being subsequently purchased, until the target size of the purchase operation was met.
Though actually yields bounced around quite a bit -- if you had been accumulating for months, QE worked out well for you, but if you bought after the announcement on October 6 and sold into the auction you probably lost money.
To contact the author on this story:
Matthew S Levine at firstname.lastname@example.org
To contact the editor on this story:
Toby Harshaw at email@example.com