Tuesday, January 6, 2009

Coping with failure


It might be imperceptable to the naked eye, but the nature of credit is changing.

The seismic changes on the Fed’s balance sheet are going to be felt in a myriad of ways: most obviously as the yield curve flattens out and effective rates begin to zero.

Most worrying for the Fed in the short term will be the danger of repo failures. The repo market is a huge, highly liquid and normally very safe forum for interbank lending. Institutions swap holdings of US Treasuries for cheap cash in “repurchase” agreements, whereby the buyback of the notes is assured at a certain date in the future.

The problem, as we have highlighted previously, is that in a zero-rate world, the incentives to “fail” on those agreements can outweigh the incentives to stick to them. The Fed, analysing the unusually high repo fails which occurred in 2003 provides a technical, but lucid, example:

Suppose a dealer sells $10 million (principal amount) of Treasury notes for settlement on Monday, August 11, but does not have the notes available for delivery that day. The dealer can either borrow the notes to make delivery or fail on its delivery obligation. For expository purposes, we assume the price of the notes is $9.98 million and the specials rate is zero.

If the dealer borrows $10 million of the notes for one day at a specials rate of zero, the dealer receives the notes from the collateral lender against payment of the current market value of the notes (which, for simplicity, we assume is also $9.98 million) and redelivers the notes to the buyer against payment of the previously agreed-upon $9.98 million price. The dealer is then obligated to return the borrowed notes to the collateral lender on August 12 against payment of $9.98 million. The balance due from the collateral lender on August 12 is the same as the amount of money borrowed on August 11 because the interest rate on the special collateral RP is zero.

Alternatively, if the dealer simply fails on its delivery obligation to the buyer, the delivery is rescheduled for the next day at an unchanged price. Thus, the dealer becomes obligated to deliver the $10 million of notes to the buyer on August 12 against payment of $9.98 million.

In both cases, the dealer has an obligation to deliver the $10 million of notes on August 12 against payment of $9.98 million. Therefore, the dealer may be indifferent between borrowing the notes on a special collateral RP at a rate of zero and failing on its delivery obligation. It follows that, in the absence of any ancillary costs or penalties, failing is preferable to lending money at a negative rate of interest.

Indeed, it can be rather profitable to fail.

Consider a trader who does not own a ten-year note but who nevertheless agrees to lend the note over the interval from July 15 to July 29, 2003, against borrowing $10 million at a zero rate of interest. Suppose the trader fails to deliver the note on the scheduled starting date. Regardless of whether the trader delivers the note late or not at all, the trader will not owe its counterparty any interest because the interest rate on the repo contract is zero. Suppose also that the specials rate on the ten-year note for RPs ending July 29 rises to 0.50 percent on July 22. The trader can then borrow the note from July 22 to July 29 against lending $10 million-thereby earning $972 interest [$972 = (7/360) x 5 0.50 percent of $10 million]-and deliver the borrowed note against its original repo contract-thereby borrowing $10 million at a zero rate of interest for the seven days remaining on that contract. The $10 million borrowing funds the trader’s loan of $10 million and the trader makes a net profit of $972.

In terms of the systemic functioning of the market, however, failures are not a good thing. They severely impair liquidity. And from the Fed’s point of view, they directly impact upon the effectiveness of monetary policy transmission into the real economy.

Which is why in November last year the Treasury Market Practices Group announced plans to institute fails charges - fines - to the tsy repo market. On Monday, the TMPG released further details of the plan:

The timeline below notes key dates identified for the implementation of the fails charge process, along with tentative dates for the dissemination of timelines and implementation guidelines for the other announced recommendations


Full details of the fails charge policy (in a handy Q&A format) are available here.

Assuming the plan works, of course, then failures should fall off, but will that make the Treasury repo market viable? There are incentives to lend at negative interest rates to be sure. As was the case in Japan during the late 90’s, and the US in 2003, repos can be conducted at negative rates (whereby a lender would pay money to lend) because, particularly in low liquidity environments, demand for Treasuries is often needed to meet other collateral delivery obligations.

The problem, though, is that in the current climate, there is rather a glut of Treasuries. So much so that there is even speculation of a Treasury bubble. The TARP continues to grow. In other words, how likely is it that banks will be in need of Treasuries - so much so that they’ll give money away to get hold of them? It seems hardly likely, particularly also when banks themselves have huge cash reserves on hand at the Fed.

So is it possible that the latest moves by the Fed and TMPG, will further kill off the interbank lending market rather than restore it to normality? As with the other Fed open market ops, will these actions simply make all institutions dependent on the central bank as the sole source of liquidity? Lender of first, last and only resort? Borrowing from - or rather, in this case, lending to, peers just isn’t attractive.

Is that perhaps, though, exactly what the Fed wants: to restore the power of its monetary policy.

No comments: