Glitches in the Bank's signals

I apologise for misleading readers last week about the likelihood of a base rate rise in September. I believed with far too much certainty that the Bank of England's operations in the money markets the previous Thursday meant that there would be no immediate rate rise, and was instantly proved wrong. So wrong, according to the letters column of this newspaper, that my remarks will be a cornerstone of the new Guinness Book of Records for embarrassing quotations.

Although I am pleased to have finally made the Guinness Book of Records, I must humbly point out that every City analyst, and the money markets themselves, all agreed that rates had been left unchanged at the monetary meeting the previous Wednesday.

Admittedly I can be accused of sour grapes, but it seems to me clear that something went badly awry with the Bank's signalling methods. Without wishing to take anything away from a good and timely base rate rise, the whole episode points to glitches in the new monetary system.

Here is a recap of events as far as we know them. The Governor recommended a base rate rise to the Chancellor on Wednesday 7 September. The Chancellor said he would need time to consider this request and closed the meeting.

Next morning the Bank of England offered assistance to the money markets in a repurchase agreement that involved unchanged dealing rates until the date of the next monetary meeting on 26 September.

The markets took this as an unmis

takable sign of unchanged policy and market interest rates fell sharply. But, at lunchtime on the Friday, the Chancellor phoned the Governor to tell him that he could now agree to a rate rise, which was implemented first thing on Monday morning. It was presented by the Government as a pre- emptive move to head off future inflation.

Various questions are raised by this episode. First, why did the Chancellor need to postpone the crucial decision for two days after the regular monetary meeting? Going into the meeting, there were two possibilities. Either the Governor would recommend a rate rise or he would not. Surely it was not too much to expect the Chancellor to have a game plan suitable for either contingency.

Decision was taken outside mechanism

A benign interpretation of the Chancellor's action is that he needed time to reflect on a key economic decision. An alternative interpretation is that he wanted to take political soundings about the advisability of going ahead - something that is not supposed to intrude into the new system.

Once the mechanism admits of the possibility that the Cabinet, backbenchers and party officials may have an input its monetary credibility is automatically undermined.

A further possibility is that the Chancellor wanted to wait until he had seen the producer and retail price statistics that would be published the following week. It transpires that he was given the retail price index at 5pm on Wednesday night, almost a week early, and he probably received the producer price numbers on Thursday.

This is important because these figures showed a noticeable up-tick in inflation. So the rate change was not pre-emptive after all. And that makes it much less impressive to the markets.

The point is that the Chancellor took the decision outside of the monetary mechanism that had been presented to the markets, based on information that no one knew he possessed.

This smacks a little too much of the old opaque system for comfort. The new system will impress only if everything is above board.

The next set of questions involves the Bank's signal, or non-signal, on the Thursday morning. Its dilemma was acute. It knew the base rate decision was in the air, but it did not want to hint to the market that this was the case. Had it done so, market panic over a Chancellor-Governor split would have ensued.

Nor, however, did it want to give an unequivocal signal that base rates had been left unchanged, since this could later be seen to have been misleading. It seems that it responded by trying to do exactly what it would have done on a normal morning, thus attempting to give no signal whatever. Unfortunately, this backfired badly.

On Thursday 8 September, the money markets were heavily short of funds to the tune of pounds 1.3bn. The Bank had to relieve the shortage, lending to the market either by buying bills outright, or by doing a repurchase agreement that in effect involves a fixed- period loan to the markets. It chose to take out virtually the whole shortage immediately in the morning, offering a repurchase agreement that expired on 26 September.

The market interpreted this as meaning that the Bank would be willing to lend at unchanged rates until the next monetary meeting - in so many words, no change in rates. An extra complication was that the previous month the Bank had calmed the markets and apparently signalled no rate change with a series of similar repurchase agreements.

No doubt the Bank would argue that its recent practice has been to offer a repurchase arrangement each time the shortage exceeds pounds 1bn, so nothing should have been read into it. But this practice is very recent and is not invariable.

Unfortunate choice of maturity date

According to Martin Brookes, my colleague at Goldman Sachs, there have been 76 days since mid-1993 on which the forecast shortage has been more than pounds 1bn. On 10 per cent of these days, including two as recently as August 1994, there has been no repurchase agreement offered, so it cannot be said that the rule would be viewed by the market as invariant. Furthermore, the average amount of the shortage taken out by these deals was only 27 per cent, and one-third of them occurred in the afternoon.

To take out virtually the whole shortage instantly in the morning was a move without precedent on any of these 76 days. This, and the decision to set the expiry date to coincide with the next meeting, seemed a highly aggressive move, and one the Bank would not have undertaken unless it intended to deliver an unmistakable message.

Was there an alternative? The signal certainly would have appeared less definitive if the Bank had removed part of the shortage by buying a few bills at unchanged rates early on Thursday morning and then offered a repurchase agreement later in the day if forced. It was the degree of aggression in the move, and the unfortunate choice of maturity date, that seemed conclusive.

The absent ingredient in this messy tale is precisely the ingredient that was intended to be central to the new system - transparency.

The main point of my remarks last week was that the Bank's signalling mechanism is reminiscent of the system of wires and pulleys by which the Sea Lords, sitting in Whitehall, used to move semaphore flags in Portsmouth. Until last week I fondly imagined that I could read the semaphore. But, with all due respect, if not one single City analyst, and no one in the money markets, can in fact decipher the code, who on earth can?

Two simple changes would solve all these problems. First, base rate decisions should be taken in the monetary meeting itself and not at some unspecified future time. Second, as soon as the meeting is over, its decisions should be announced to the world in plain and simple English, which even I can understand.