In the real world the cost of borrowing gold outright is never negative – no bullion bank will pay you to take gold. In fact, I am aware that some lenders have a minimum rate below which they will not lease. Makes sense, would you risk lending 1 tonne of gold worth $37 million on an unsecured basis at 0.1% for 3 months just to earn $9,400?
So why does the LBMA report negative lease rates? Our starting point is how the lease rate is calculated:
Lease Rate = LIBOR – GOFO (see the LBMA’s Guide for why this is so)
First point to note is that the lease rate is calculated from LIBOR and GOFO; the LBMA does not question its market making members for their actual lease rates. It is therefore based on the accuracy of LIBOR and GOFO. If we look at how these two rates are determined (see here and here) then we see a number of differences:
1. Set at different times – GOFO rate submitted at 10:30am and fixed at 11am, but LIBOR rates are requested between 11.00am and 11.20am and fixed shortly thereafter.
2. Set on different sides – LBMA’s website GOFO is the rate “at which the Market Making Members will LEND gold on swap against US dollars”, which involves using the USD interest bid rate. For LIBOR banks are asked “At what rate could you BORROW funds, were you to do so by asking for and then accepting inter-bank offers in a reasonable market size”, that is the USD interest offer rate.
3. Set by different banks – LIBOR is set by 16 banks, GOFO by 8, with 6 common to both. LIBOR drops the top and bottom quartiles before averaging, GOFO drops the highest and lowest before averaging. It could therefore be possible that the 4 banks use to set LBMA’s GOFO (which they would calculate/relate to their estimate of LIBOR) don’t have their LIBOR rates included in BBA’s LIBOR. Probably worth noting that within a bank LIBOR and GOFO would be set by different desks.
Now these are minor differences as we would not expect rates to move too much between 10:30am and 11:00am, or much difference in the bid/offer spread, or too much divergence between banks on their rate so the dropping of high and low rates should not affect the average too much.
However, I think when rates get close to zero, these differences could have a material impact. Consider also that questions have been raised about LIBOR’s usefulness at these low rates, see here and here.
The fact is that GOFO and LIBOR are not “in alignment”. The resulting calculated lease rate is therefore just an approximation based on two averages. Caution should thus be exercised when trying to draw conclusions from it.
GOFO, however, should be able to be relied on. It should relate to the basis, although not equate to the basis as the basis is calculated from futures prices whereas GOFO is a forward rate – the economics of those two are slightly different.