IS NEGATIVE GOFO SIGNIFICANT? (Now With A Backwardation Chart Addendum)

S. Roche alerted STFU comments’ readers last Monday, 8 July (http://screwtapefiles.blogspot.com.au/2013/07/figured-it-out.html#more) that ‘GOFO negative out to 3 months, this has meant fireworks in the past!!’ and later that it had become negative through six months.  That first observation engendered a flurry of related comments, but why?.  What is the significance of that phenomenon?  Well, to begin with, it is rare, having only occurred thrice in recorded history before this, in 1999, 2001, and 2008, and then only very transiently.  In all three instances, the price of gold went higher, sharply but very briefly in 1999, and less steeply but more protractedly on the later two occasions..

So, the still burning questions are: ‘Will history repeat (or at least rhyme)?’  Or, ‘Is it different this time?’

For those who are not familiar with GOFO (Gold Offered Forward Rate) published by the LMBA on a daily basis (http://www.lbma.org.uk/pages/?page_id=55&title=gold_forwards) and the related Gold Lease Rate (GLR) published with a week’s delay, a good place to begin is Bron’s 2008 critique (http://goldchat.blogspot.com.au/2008/10/misinterpretation-of-gold-lease-rates.html) of Brian Kelly’s SeekingAlpha article (http://seekingalpha.com/article/100677-misinterpretation-of-gold-lease-rates-and-why-gold-could-rise?).

For those desiring only a quick refresher, GOFO is related to GLR via LIBOR (London Interbank Offered Rate) according to the formula GOFO = LIBOR - GLR, so when GLR is greater than LIBOR, GOFO is negative, the corollary being that when GOFO is greater than LIBOR, GLR is negative.  Contra the naïve assumptions of the less well-informed commenters out there on the gold blogosphere, a negative GLR does not mean that folks will pay you to borrow their gold, just that it costs less to borrow than dollars.  According to Bron (above), GOFO is the “amount that can be earned from the gold carry trade”, which amounts to selling gold to buy dollars when the latter yields higher interest.  So, when the GLR is higher than LIBOR and GOFO is negative, it simply means that it costs more to borrow gold than dollars.

Negative GOFO rates do not necessarily mean that they are in backwardation over the time curve, nor does it necessarily mean that gold futures are in backwardation, as James Turk and some others seem to assume, as can be seen below.

S. Roche’s breaking news in an emailon 8 July (1M - 0.065% 2 & 3M Neg as well!) engendered a lively email exchange among a few of us down under here in the Land of Oz, which included informed comments and many of his usual excellent charts from Nick Laird (sharelynx.com), insights from industry-insider Bron (goldchat.blogspot.com.au), insightful observations from trader Roche, and a few questions and opinions from yours truly, Slow Loris Larry.

In at least approximate chronological order, here are the highlights of our email exchanges (condensed and/or slightly edited).  Links to various articles appearing on various sites that we discussed are included.  Roche posted this link to the first internet article on the subject on STFU comments on 9/12 (URL above in first paragraph) http://www.zerohedge.com/news/2013-07-08/historic-inversion-gold-gofo-rates-turn-negative-first-time-lehman

9/7 Our relevant email exchanges got going when Bron sent us the ZH URL above:

A surprisingly qualified post by ZH.  Usually they just go with whatever single explanation makes for the best headline.

Then Nick sent us his first graph::

Here's a chart I run on the GOFO's

They have been below zero back in May

                                                        All charts can be clicked to enlarge.

And Roche’s reply is:

I saw that on May 22nd and I think it is a typo because of May 21 and 23, they don't correspond to negative 1M on May 22. I have emailed the elves at the LBMA last night and I will let you know.

(Much later, on 15/7, he reports: May 22nd has been adjusted down twice on the LBMA website since I emailed them but is still showing negative.)

Nick, again:

Here's some better charts using the LBMA data

GOFO since 2005

And Lease Rates for the same time

                                                         All charts can be clicked to enlarge.

SLL then chimed in:

Let's see if I have this right: 

GLR (Gold Lease Rate) = LIBOR - GOFO (Gold Forward Rate)
Also GOFO = LIBOR - GLR,  so IF GOFO < 0, GLR > LIBOR, meaning that it costs more to borrow gold than dollars.
James Turk, blogging on KNW (http://kingworldnews.com/kingworldnews/KWN_DailyWeb/Entries/2013/7/11_Turk_-_We_Are_Witnessing_Historic_%26_Shocking_Events_In_Gold.html), says that the LMBA website is now showing gold in backwardation because of the negative GOFO.
Is that a necessary relationship?  Dear old  Prof. Fekete taught me that Backwardation is the situation in which the spot price is higher than futures prices (but to be meaningful it should extend out more than a month or so and be consistently proportional along the price curve), the cause being that arbitragers will no longer sell for present (or near) delivery and buy for later delivery because they won't accept the risk of not getting eventual delivery (as Turk also points out).
Turk goes on to say that SIFO (the Silver Forward Rate), which has not been reported for the last several months, was in Contango (futures prices being higher than spot prices) BUT no one could trade at those rates, implying that silver might have actually been in backwardation since around that time.
So, does this make the drawdown in gold from both the COMEX and GLD but not silver from the COMEX and SLV all the more mysterious?

And Nick replied:
Here's gold's current spread

 

                                              All charts can be clicked to enlarge     

You can also find GOFO rates here

http://www.sharelynx.com/gold/LBMArates.php

Negative GOFO and Backwardation.  Note that Nick’s chart does show gold in backwardation, but for four months, only, with August being the deepest.  SLL

In a following message, Nick adds:.

I've long thought that since they butchered the Libor rates that the Gold Lease Rates shown officially were bunkum.

And that's been since 2008 when the US Gov't drove their rates down to zero....

Roche replies:

I agree with Nick and think GOFO stands alone, it is easier to understand that way too rather than cloud the issue with Lease Rates which don't apply to any transactions and are just shorthand for the cost difference between gold and USD.

Sorry to bang on about this but I am in the camp that GOFO is about collateral and collateral chains, at the moment being unwound. It is the only thing that makes sense and justifies the huge daily volume in gold trading. Gold serving as collateral in the financial system is one of its main functions and can drive markets at times, in a big way. Imao. 

Bron then clarifies ‘real GLRs’

“Lease Rates which don't apply to any transactions.”

Perth Mint does lease metal outright, so Lease Rate is real (and never negative) and there is a leasing (borrowing and lending) market just as there is a forward GOFO market. Which drives which is hard to say and probably changes over time.

Roche then asks:

Do you use the LBMA lease rate or your own OTC?

And Bron replies:

All lease rates are done OTC and depend on who you are (your credit rating). GOFO, LIBOR and GLR are idealised general/mean rates and don’t directly relate to what a counterparty will get.

On 10/7, Nick sent a link to Gold GOFO/Libor/Lease Rates

                                                         All charts can be clicked to enlarge

And Roche says:

Jack Farchy from FT chimes in http://www.ft.com/intl/cms/s/0/bd819998-e8b2-11e2-8e9e-00144feabdc0.html

On 11/7, SLL posed and then (sort of) answered a question:

I still have a basic question that hopefully has a simple answer:
Although the two phenomena are clearly linked to some degree, I am wondering if negative GOFO invariably also means backwardation in the futures/forwards exchanges/markets, and vice versa?
I think that the answer is 'not necessarily’, and that there could be one or the other alone without the necessity of the corresponding one also being evident.

Roche answered:

My understanding is that gold forwards are in backwardation.

Bron also answered:

GOFO is like a secured forward and as such is a similar financial instrument to a futures contract. Futures are "secured" by the requirement to deposit margin. That can result in different economics (profit) than a forward or gold swap. 

The "basis" is more just the revenue part of the transaction and does not take into account the funding/opportunity cost of the futures margin or other costs, so the basis percent may come down to be closer to GOFO.

NB:  ‘Basis’ refers to the price curve of futures/forwards prices versus the ‘spot price’ however that may be established.  If the basis is positive, a market is in contango.  If the basis is negative, a market is in backwardation.  SLL

Roche announces:

And Izabella Kaminska chimes in: http://ftalphaville.ft.com/2013/07/10/1561322/the-hand-of-gofo-strikes-again/

SLL responds:

Here are a couple of comments on Kaminska's post:
1) I think you have been right in saying that what is going on is related to financialisation of gold, as K implies (as I understand, or may misunderstand, what she is saying).
2) I had real difficulty understanding some of it, even after I followed all of the links and then reread the piece.  For instance, she says:
"To entice cash lending against gold, either lease rates had to go negative or the GOFO lending rate had to take the slack by turning negative."
Since GLR is a reciprocal of GOFO through their inverse relationship to LIBOR, negative GLR and negative GOFO are opposites.  So why/how do those two opposite situations have the SAME effect on willingness to lend against gold?

SLL then announces:

Dave in Denver also posts on GOFO, http://truthingold.blogspot.com.au/2013/07/gofo-explained-and-why-its-now-very.html#comment-form

Bron also announces:

Another good article http://goldnews.bullionvault.com/gold-borrowing-071020136

12/7 SLL informs the group of the following:

Steve Saville, another Aussie, who produces The Speculative Investor, wrote in his Interim Update 10 July 2013 on the subject, stating ‘This means that a small shift into backwardation in the current interest rate environment will be far less significant than a small shift into backwardation in an environment where the 3-month T-Bill has a more normal yield of, say, 3%-5%’.

So, Saville definitely thinks ‘this time is different’!

12/7  Roche responds to Saville’s conclusion:

That's the only proviso that I've got in mind. But, I don't think the short interest was as extreme last time so maybe it will balance out.

Bottom line, no one knows: last two times there was a pull back after the initial surge so I am trading with that in mind as well.

I am betting that history rhymes. Cheap money has only meant more leverage to get the same bang for the buck.

12/7 Nick presents a great chart and the final words to date to go with it:

Here's a long term view of all the Gold Rates.

                                                         All charts can be clicked to enlarge

With regards the GOFO rates - they have not inverted at all - their time/premium is still normal.  What they have done is gone into negative territory.

Interestingly enough one can see back in the 1990's where the GOFO did invert, i.e. longer term rates fell below short term rates.  To clarify my thinking on backwardation & contango:  With rates series there is an interest rate due (say Libor) for each & every day.  Thus a 3 month rate carries a lower fee (interest rate) than a 12 month rate.  Thus as you look out along the rates curve what you are seeing is the time premium. 

A normal rate curve - one which is in contango, shows the time price of the rate rising.  A rate curve in backwardation shows the situation whereby the rate curve inverts & you get the 12 month rate cheaper than the 3 month rate.

This disparity shows the time price premium of interest no longer exists due to demand structures which have inverted the curve.  So to show true backwardation the rate curve needs to invert - for the time/price rate to invert & thus show demand overtaking the rate structure.

As such with the current gold situation whereby the cash price is slightly above the active near month (August) & the next near active month (October) we are seeing slight backwardation in these near months.  December has not yet inverted.

I will soon build a chart to show the spread of the futures contracts & their contango/backwardation.

SSL concludes this post:  Won’t it be interesting to see what time will tell us?

ADDENDUM

Following are Nick Laird’s Sharelynx.com charts for Gold Futures Contract Spreads for Jul-19 and Jul-22.

                                            Click to enlarge charts.

                                                         
In a comment posted below dated Sun July 21, I reported that there was increased and prolonged backwardation on the July 19 chart (now shown at the top, above) in comparison with the one originally posted for July 8 (shown below), but I could not get the Jul-19 chart to display in that comment).

Nick has now updated that chart to July 22  (the middle one, above), and it shows that backwardation had then all but disappeared.

GOFO rates for July 23are still negative for the first three months in about the middle of the range since July 8.

The cost to access Nick's huge number of continually updated Sharelynx charts for a year is extremely modest.  If you have gained anything from those that I have posted here, you might consider subscribing.

Note, however, that charts that update on a daily basis are replaced, and that old ones are not normally accessable unless you have saved them locally..

SLL