Friday, March 21, 2008

Financial Markets Getting Really Weird: Negative Nominal Interest Rates

Textbooks and most theoretical papers tell us that zero is the floor on nominal interest rates. But news reports in the last few days, noted by both Paul Krugman and Brad DeLong, have announced that some repurchase agreements in New York have had negative nominal interest rates attached to them, parties paying to lend money to others. This seems to be tied to very low short term rates on US Treasury bills, barely above zero, pushed down by TED risk spread so that one month bill rates are about 2% below the 2.25 overnight federal funds rate, which is private interbank loans. Negative nominal rates were last seen in the US on repos when the ffr was at 1%, especially during August to November, 2003, as reported in a paper by Michael J. Fleming and Kenneth Garbade in the April, 2004 Current Issues in Economics and Finance, of the New York Federal Reserve Bank.

It is not just the occasional repos that have had negative nominal interest rates. Barrons in 2006 has reported on a curious financial instrument issued by Berkshire-Hathaway known as "squarz." These have had negative nominal rates on them. Also, in 1998 for a brief period very short term government securities in Japan had negative nominal interest rates, during the pit of their deflation, as reported in the Monetary Trends column of Daniel Thornton from the St. Louis Fed in January, 1999. And, finally, although this has never been reported in print, I was personally told by the individual who handled dealings between the Fed and Freddie Mac that on Dec. 31, 1986, the last day of the old tax code before the Reagan simplification, when many were trying to close a lot of deals, the federal funds rate itself briefly went into negative territory down to about - 1/2 percent, although it also soared as high as about 18% on that rather wild ride of a day. In any case, things are getting weirder and weirder in the current frenzy of the financial markets.


Sandwichman said...

Presumably the parties paying to lend money to others are lending other people's money at negative interest rates?

Unknown said...

somewhere, Silvio Gesell is smiling.

juan said...

TYhanks Barkley,

From the Fleming & Garbade paper (2004):

The 2003 episode is also interesting because of the specific circumstances that led to negative interest rates. The option of Treasury market participants to fail on, or postpone, delivery obligations with no explicit penalty usually puts a floor of zero on repo rates. In 2003, however, ancillary costs of failing increased as settlement problems in the May ten-year note persisted. The increased costs ultimately led some participants to agree to negative interest rates on RPs that provided the May note as collateral.

Or, given settlement problems, negative nominal interest was the least expensive option.

Other hand, 3/20/08:

“As soon as we approach zero we have this somewhat obscure effect whereby it’s cheaper to fail than to honor a negative interest rate,” said Matt Thomas, head of repo trade at BNP Paribas in New York. In such a scenario, the recipient of the security would be motivated to hold onto the Treasury note because swapping it back for the lent-out cash would end up costing them.

But Scott Skyrm, head of the repo desk at brokerage Newedge, noted that a pickup in fails Wednesday and said that could further down the line indicate disruptions to the settlement system and securities being permanently removed from the market.

“We can go either way,” he said in research.


Though I tend to think of such moments as another manifestation of the larger tension between real and financial, 'weirder and weirder' is more apt.