Select Committee on Public Accounts Minutes of Evidence

Examination of Witnesses (Questions 40-59)



  40. I notice one of the NAO's expert panel cites the work of two Nobel Prize winners—Friedman and Miller—and basically says that they made a mistake in their analysis of how auctions operate. You must take some satisfaction for out-guessing Nobel Laureates?
  (Mr O'Donnell) I think Professor Binmore should take the satisfaction, to be perfectly fair. We are in this country really lucky in having some brilliant auction theorists and I think Ken Binmore and Paul Klemperer, for example, have done the public sector an enormous service in terms of developing auction theory which has been used to incredible effect, not just here in telling us things about the uniform price versus descending price auctions, but indeed in mobile phone auctions, where they were both heavily involved.

  41. There do seem to be critical comments in the Report about the lack of modelling for predictability of the sales. How do you take those on board?
  (Mr O'Donnell) This is a very esoteric subject, in the sense that trying to model precisely how much one gains by predictability is difficult. There is some academic work cited by Ken Binmore, I think on page 44 he refers to some work there, but it is true that we cannot easily come up with an answer as to how much one gains from predictability. Certainly market participants have told us they value having auctions at a predictable time, that that is to them important, so they can simply organise their books ahead of time, they can make sure the liquidity is there and, mundane things, like making sure their traders are there and the people who are auction experts.

  42. What is your advice to the countries which have decided to maintain their high gold stocks, like France?
  (Mr O'Donnell) I think that is a matter for them. I think each country needs to look at its own preferences with regard to the riskiness they want to have in their portfolio and they will obviously come to views about what they think are perceptions about long-term prices. All we can do is make transparent the analysis we have gone through and it is up to them what they do.

  43. Do you see future stability lying in that balance between gold, the euro, the dollar and the yen?
  (Mr O'Donnell) I think you will find that when countries consider their overall reserve holdings, those for now at least and for the next few years will be the big currencies, but if you look 10 to 20, maybe even 30, years further on, you have to say some other things could develop. For example, we know that at current growth rates the Chinese economy will become one of the biggest in the world, so that may well have implications as to what one considers as the Asian part of that portfolio.

Mr Davies

  44. Good afternoon. Can I say, first of all, I completely agree with the strategy of diversifying the portfolio for reasons of interest and capital values, but I am somewhat concerned about the process in terms of maximising value for money. It is clear from what has happened in Figure 15 that on 17th May there was an announcement to sell off a very large proportion of our assets and by 6th July it was apparent to your market the price had dropped quite dramatically. Was there any option there to reduce the amount we were going to sell?
  (Mr O'Donnell) The markets will respond to the news they are given. If you are saying, should we have announced a smaller proportion of gold sales, that might have had a different impact on the price, but then we would not have achieved our objective of changing the portfolio in the way we wanted to to reduce the risk to the desired level. So we would have ended up with a portfolio with a higher proportion of gold if we had not sold as much, and it would have been much riskier. We thought it was better to decide we would move to the point which we thought was right in terms of the risk portfolio, announce that in advance and give the market some certainty that that is where we would stop.

  45. But the issue there is timing, is it not? Are you in a position to announce you are going to sell some gold subject to a reserve price, for instance?
  (Mr O'Donnell) You can say that but that creates quite a lot of uncertainty in the market because if you pre-announce the reserve price then in a sense you are putting in a floor.

  46. And there is a natural tendency for the market to go to that reserve price?
  (Mr O'Donnell) Exactly.

  47. Not below it though?
  (Dr Mills) In terms of when we hold an auction we do not announce a price for that reason, that it gives potentially strategic bidders a target to get the price down to. So we have a reserve price below which we will not sell but we do not announce it for that reason. If we were going to start the programme and say we are going to sell 415 tonnes so long as gold does not fall below x dollars per ounce, in one sense that would have stabilised the market but I do not know what the price of gold should be other than the current market price. Hence if the market falls or rises (because I generally believe that markets take into account all available information) then that is the fair price for gold.

  48. It seems to me that if I were managing a portfolio I would have a number of stocks and I would have some sort of rule of thumb that would say I will sell this stock after it reaches a certain price and below a certain price I will not sell it or buy it. What is happening here is instead of doing that we are saying we will sell it anyway, the market adjusts to the price and we might be lucky, the price might go up or it might go down, but it seems a bit strange compared to what a fund manager would normally do.
  (Mr O'Donnell) Fund managers are operating to relative short timescales for their greater return. For us we are looking at a long-term position. We will not want to long term—

  49. That being said, you said a minute ago you wanted to rush in and get this amount sold off to re-balance your portfolio within a reasonably tight timescale. Having a long-term managed approach to whatever it is, 50 per cent in gold, having less than that, whatever the figures are—
  (Mr O'Donnell) From about 44 to 20.

  50. That is fine. You are saying we will do it by this time and that creates market expectations so it is easy for the market to make money as opposed to you, is it not?
  (Mr O'Donnell) I do not think so. We give them predictabilities in saying we are going to sell over this long period, so over that long period there are all sorts of factors which will affect the price, sometimes it will have gone up a bit, sometimes down a bit, and basically we are getting the average price over the whole period of the sale. Again it is another way of spreading the risk.

  51. Tell me about the currency configuration that you chose. You chose 40 per cent euro, 20 per cent yen and 40 per cent dollar. Obviously the yen has been quite unstable in recent times and the euro has been gradually dipping, the dollar has been taking off and it has slowed down, and therefore you have got a balance of up and down to a certain extent. Is there any consideration about non-currency holdings and, indeed, having a very different balance between these?
  (Mr O'Donnell) Since we do not regard ourselves as having any special knowledge of where these currencies will move over the long term, and this is a long-term decision as to what to hold our currencies in, we take a long-term view that it will be 40:40:20. We did this some time ago, pre the announcement on gold sales and we simply stuck with that.

  52. You said you reached the decision a long time ago. How long from the South East Asian collapse did you decide to have 20 per cent yen?
  (Mr O'Donnell) The decision on 40:40:20 was first announced in January 2000.
  (Dr Mills) But it was in the accounts in the exchange equalisation account for 1996-97.
  (Mr O'Donnell) This has been a long-term view.

  53. Can I take you on to something else. Obviously what has happened is after your second auction the price lifted and then fell before your third auction as a result of co-ordinated action across Europe to gradually release gold in a controlled way. In hindsight if you had done that from the start, if we had had a more pro-Euro approach, sat round and said we have got the same issue here, we are all heavily concentrated in gold and we want to balance our portfolios let us gradually release it together, would that not have been a better deal for Britain?
  (Mr O'Donnell) The way it worked out was rather a good deal for Britain in that we pre-announced our sales and then the European central banks came together to announce an overall limit, but it was accepted that our sales because we had pre-announced them would be there. The rest of them in a sense had to fit round us.

  54. Looking at Figure 15, in simple terms it seems to me that the average price drawing a simple line across this scatter was something like 275 and then the average price after the second auction when the price lifted was nearer 285, so presumably the Europeans selling after that time were achieving a higher average price?
  (Mr O'Donnell) No.
  (Dr Mills) The European agreement was limiting sales to what had been pre-announced so it included the Dutch sales, the Swiss sales and our own and all the other banks who had not been included in that overall limit were essentially tying their own hands and saying we are not going to sell any more gold than has been pre-announced for five years. Essentially no other central banks apart from those covered by the agreement were selling, so anyone not mentioned or not included in that overall total was not taking advantage of that.

  55. You have got 300 tonnes or so left of gold?
  (Mr O'Donnell) We will have by the end of the programme.

  56. But there is more stability, I think I am right in saying, in the gold price now. Presumably you will continue to release at a slower rate gold to re-balance your portfolio? Is there a rule of thumb that you will not sell below a certain price threshold or can you not do that? Are you still at this thing we are going to sell 100 tonnes of gold whatever the price in three weeks' time and then it comes along and suddenly the price drops and then the next day the price goes up? You cannot guard against that?
  (Mr O'Donnell) That is not what has happened with auctions. I would be surprised if that is what did happen because then there would be opportunity for arbitrageurs to make money by simply moving across that. That is not what has happened because the market is quite a good market and looking forward I would expect us to take whatever market prices are. We do not try and out-guess the market in that sense.

  57. When you announced on 7 May that you were going to sell off this gold, you presumably predicted the price would fall?
  (Mr O'Donnell) Yes.

  58. Is it reasonable to assume that if you had announced half the sale you did announce that you could have predicted that the price would fall less?
  (Mr Plenderleith) Not at all because the difficulty was if the Government had been able to say, instead of it was going to sell 400 tonnes, it was only going to sell 200 tonnes and that was the end of it, then yes the price effect would have been less, but that was not the Government's intention; the Government's intention was to sell the 400 tonnes. It needed therefore to make that clear from the outset because if it had not stated an amount the market would have assumed it was going to sell all 700 tonnes and the price impact would have been much worse. It was advantageous to the Government to state the amount at the outset.

  59. If it had not announced 400 tonnes but announced 200 tonnes with a view to another 200 tonnes, but not told the markets, would the price have dropped less?
  (Mr Plenderleith) Much more because the market would have feared the worst and assumed that beyond that initial 200 there was a further 500.

previous page contents next page

House of Commons home page Parliament home page House of Lords home page search page enquiries index

© Parliamentary copyright 2001
Prepared 19 September 2001