To roll or not to roll, that is the central bank's question

Yesterday I was dismissive of the recall of Hong Kong's gold as significant, but it is another bit of evidence of a shift in central bank attitudes towards gold. Far more significant indicators include (see this MineWeb article):

* China's announcement that it had moved 454 tonnes of gold into its reserves since 2003
* Central Bank Gold Agreement (CBGA) quota being reduced from 500t to 400t a year
* Russia's Prime Minister stating that it should hold 10% of its reserve assets in gold

It points to a renewed appreciation of the role of gold in turbulent times. Recalls of gold like Hong Kong may also indicate a reassessment of counterparty risk. Moves to return gold are eminently sensible, of course: what is the point of a country having its gold out of its immediate physical control if everything goes to hell. That is really the whole point of having gold reserves. In a time of war (not that I'm suggesting that is where we are heading) you ain't going to be able to buy guns or food from another country with your funny paper money.

Some have claimed that repatriation of gold by other central bankers following Hong Kong's lead will translate into higher gold prices. However, this depends on the extent to which that gold is actually sitting in a vault somewhere or has been lent out to bullion banks. If the former, then obviously there is no effect on the price – the gold is just changing location. If the latter, then it could be potentially explosive if Frank Veneroso's estimates of leased gold of between 10,000 and 16,000 tonnes are correct.

I would point out that central banks can't just recall gold mid-lease, they have to wait till it's maturity. Consider also that the leases will have been made over varying terms, from a few months to a few years, and all at different points in time. This means that all of the central bank leases will mature over a number of years. What the term to maturity of this global lease book is, is hard to say. I'll have a stab at most of it being 1 to 2 year leases, but am prepared to stand corrected.

So not all of Mr Veneroso's leases will be recalled immediately, or to be more accurate, declined to be rolled. Plus not all central banks will decline to roll their leases (although that may change depending on how bad things get).

Also, don't fall into the trap of assuming that all of this leased gold has to be bought back from the market to repay the gold loans. This sort of simplistic analysis is based on an ignorant view that “leasing = bad”. The reality is a bit more complex. To explain, I am going to have to be a hypocrite and be simplistic myself. There are three things someone can do with borrowed gold:

1) Manufacture it into jewellery, coins or bars. Sell these for cash. Use cash to buy replacement gold. Hopefully have left over cash = profit. Repeat many times.
2) Sell the gold. Use the cash to build a mine. Extract the gold from the ground. Repay your gold loan. Hopefully have left over gold. Sell this for cash = profit.
3) Sell the gold. Invest the cash to earn interest. Hopefully gold price drops. Use part of your cash to buy gold. Repay your gold loan. Left over cash = profit.

All of the above are ultimately promises to repay gold, but not all of these have the same risk profile. I've ranked them in terms of risk and the first two are materially different to the third. In the first two the gold loan is backed by gold, either in inventory or below the ground.

In the current gold market, one would have to consider the risk of failure low for the coin/bar business – everyone wants the stuff – and I'm sure that central banks, through bullion banks, would not consider these leases high risk and necessitating recall. For jewellery, the increasing gold price equals less sales, so we could expect some business failures, so while these leases are backed by physical it would have to be considered at some risk.

For miners it is a bit more risky. Sure they have it in the ground, but lets not forget Bre-X or Sons of Gwalia. As long as any hedging is modest and loan maturities tied to production, these would also be considered lower risk by central banks.

In the case of the first two it ultimately comes down to the extent that the lease is secured: the first two are not risk free - business ventures do not always turn out as expected. To the extent that they are not secured in some way, central banks would have to be nervous, but not as much as our third category.

In the case of the short sellers, the gold is gone and only cash is left. To the extent that a miner has excessively hedged (did I hear someone say Barrick?), then they are also in this category. The crux of the issue is to what extent have the short sellers put up collateral and more importantly, have the ability to put up more (or the willingness to put up more)?

This collateral issue I will discuss in my next post. My point for the moment is to not get awe struck by the 16,000t figure (or whatever other figure is bandied about) and think it is all going to have to be bought back, and now, and therefore the gold price is going to the moon.

If central bank reassessment of counterparty risk results in requests for leases to be repaid, then it will occur over a number of years as those leases mature. This will manifest itself as a steady stream of short covers, not as a big bang, and be a source of solid "base" demand for gold for a number of years.