It has been a while since we updated our basis analysis of gold and silver and this is primarily due to the fact that the basis wasn’t doing much of anything. It still isn’t doing all that much but there are signs of change on the horizon so we thought it would be a good time to provide an updated reference point for our ongoing commentary, which is about to go into high gear given the current market situation.
As many of you know, “basis” refers to the spread between prices in two different markets or locations for the same underlying commodity or security. These two markets may be separated by physical distance, which is how basis is measured in the grains, or by manner of ownership, which is one of the ways that basis can be measured for gold and silver. In particular, the classic measure of the basis in gold and silver involves analyzing the spread between the spot price of bullion, which is derived primarily from trading on the London Bullion Market (LBM), and the futures price of bullion, which is derived primarily from trading on the COMEX.
Although the LBM and COMEX are separated by physical distance, this isn’t the main factor creating the differences in prices from which the bullion basis arises. Rather, the main factor is the time lapse between the trade and settlement. On the LBM, physical trades settle in 2 days whereas on the COMEX physical delivery occurs at some point in the future based on the terms of the particular futures contract. For practical purposes, the basis calculation always uses the active futures contract, which currently for gold is February 2011 and silver is March 2011. In any case, the difference in prices between the spot and futures markets is due to the passage of time, and time is money. Specifically, bullion does not generate interest as does invested cash. Acquiring bullion on the spot market today instead of waiting for future delivery on the COMEX requires an expenditure of cash today and therefore interest is foregone in comparison to acquiring gold for future delivery (which requires a much smaller present cash outlay in the form of margin).
Thus we should not be surprised that gold for future delivery should cost more than gold for immediate delivery especially if interest rates are moderate. The condition where the future price of bullion exceeds the spot price is called contango and represents the normal condition of the bullion market. By “normal”, we mean that gold and silver should trade in contango as long as all other things are equal. But what if other things are not equal? For example, what if there is a rise in default risk associated with delivery of bullion in the future? In such a case, buyers may have a strong preference for obtaining gold today even if it costs more than gold promised for delivery in the future. In effect, the foregone interest can become irrelevant compared to the risk of default on the futures contract. Under such conditions the spot price of bullion can be higher than the futures price, and not only for an instant here and there but for a sustained period of time. This is called backwardation. Importantly, backwardation is more meaningful when interest rates are high and rising instead of low and falling since in the former case it is financially more painful to forgo the interest in exchange for the security of immediate bullion possession.
Unfortunately, there are some very serious problems with constructing basis studies from publicly-available price data. The futures price is not a problem as the COMEX exchange provides very detailed trade data. Discrete data on spot prices, however, is very difficult to acquire if you are not a bullion bank — for the simple reason that the spot market represents over-the-counter trading between private parties and does not have its own exchange or reporting mechanism. The only “official” data on spot market prices are the London gold and silver fixings that occur twice a day for gold and once for silver.
There is also an issue with the type of price being used since one can have a bid price, an ask price and a trade price. Despite some strong arguments made by basis proponents that the bid-ask spread is a relevant consideration to undertake in studying the gold and silver basis, we believe that using bid-ask data is fraught with problems. For example, a simple bid or ask price does not transmit information about volume and can provide deceptive results under various market conditions (as has been the case several times since 2008). We therefore prefer using actual trade data in our basis analysis. To our knowledge, the only proxy for spot gold and silver trading that exists is the FOREX market where money center bank currency desks make a cash market for gold and silver under the symbols XAU and XAG.
Another proxy market for gold and silver that has existed for a few years now is the ETF. The major gold ETF, GLD, and its silver cousin, SLV, contain pricing mechanisms that closely track spot bullion prices (for the most part). This mechanism relies on a linkage to bullion trading that takes place on the LBM. Yet the ETF price is never exactly the same as the bullion spot price, primarily because of demand and supply factors that create net asset value (NAV) premiums and discounts. For example, Jim Cramer may recommend to his flock that GLD is a good buy at the moment and that may create instantaneous demand for the ETF that is not matched in the spot gold market. In any case, it is possible to track the NAV premiums and discounts over time and this can reveal similar information about the underlying gold and silver market as does the classic gold and silver basis. For example, both the classic gold basis and the “GLD basis” can reveal large fund flows in and out of the metals that might otherwise not be apparent to market observers. Therefore we use both methods in our basis studies.
Much more can be said about the gold and silver basis but let’s go ahead and look at a few charts and save further exposition for a future date. First up is the classic gold basis. Click on all charts to open full-size versions.
You will want to avoid the chart action in late November during Thanksgiving and also ignore the gentle downsloping trend (an artifact of the fixed calculation required in order to render the chart). Instead, we are looking for major movements in the basis toward or into backwardation (the red line represents the threshold between contango and backwardation) or some other major shift. Other than single data points, which represent anomalies that appear quite often in basis studies, we can see that the gold basis has been rather well-behaved even as gold has fluctuated around record levels.
The internet may be full of dubious claims about “commercial signal failures”, imminent supply shocks and various end-of-world prognostications but the basis reveals gold to be in a normal bull market at the moment. By “normal bull market”, we mean one where prices are determined by the incremental excess of demand over supply. In other words, there is no ongoing massive rush into gold that will momentarily drive the price to the stratosphere. Indeed, the gold basis reveals that physical demand remains well below historic levels such as during 1979-1980 when gold often traded in backwardation even with interest rates at very high levels (see above for why this is relevant).
Note that an encouraging sign can be found for gold bulls at the right end of the above chart, which shows that contango in gold may have built a slight upward bias during the recent pullback but does not (yet) indicate the type of physical dumping that has characterized the major gold corrections of the past. Physical dumping tends to cause the basis to rise, resulting in higher contango, because large sales in the spot market are more difficult to absorb than in futures markets or the ETFs. The simple reason for this is that the spot market requires a dollar-for-dollar offset in physical demand whereas the paper markets can also absorb supply through short covering.
Next up we have the GLD basis chart featuring the gold ETF. In this chart the GLD’s NAV is par at the chart value of zero so there is a NAV premium above the red line and a NAV discount below the red line.
For at least the past couple of years, both GLD and the silver ETF SLV have typically traded at a very small NAV premium simply because demand has consistently exceeded supply. The excess demand, however, is only sufficient on distinct occasions to generate a NAV premium large enough to justify delivering additional bullion to the ETF. Delivery of this bullion results in the issuance of additional ETF shares, which are used to satisfy the incremental demand. On the flip side, supply has rarely been sufficiently excessive to cause the redemption of ETF shares and a consequent decline in the ETF’s bullion holdings. Indeed, one of the hallmarks of this bull market in gold and silver appears to be the maintenance of an ongoing slight NAV premium in the bullion ETFs and the consequent and constant accumulation of ETF metal holdings. We find it difficult to imagine a scenario where gold and silver could enter a longer-term correction or even an intermediate bear market absent a radical change in the ETF basis from perennial NAV premium to a NAV discount.
Moreover, we find that the ETF basis can help us evaluate the character of an ongoing price rally or pullback in the short to intermediate term. In the above chart, for example, we can definitely note that the NAV premium has declined and even flirted with a discount but still remains positive even after two days of very sharp gold price declines. Continuing to observe ETF basis action over the next several days should reveal strong clues about the nature of the current decline. One possibility is that bargain hunters may enter the market en masse around a certain price level and this would obviously have positive implications. In our experience, the basis can often provide advance notice of bargain-hunting activity. One way this could manifest is a sharp upturn in the NAV premium.
Next up is silver and because it largely confirms the story we are seeing in the gold basis, we’ll only provide abbreviated comments. First up is the classic silver basis.
Quite a similar story here as the gold basis except that we are seeing a more aggressive jump in the contango, which would be consistent with the steeper fall in the silver price and perhaps more physical selling compared to gold. Given silver’s relative outperformance of gold to the upside, none of this is surprising.
The SLV basis chart follows.
Interestingly, the NAV premium on the SLV is actually stronger than GLD and indeed I just checked my live version of this chart and there was a further uptick in the afternoon. This is again an encouraging sign pointing to the weakness in silver being temporary . . . although the ETF basis can notoriously turn on a dime and therefore no trader should take comfort from a fledgling signal like this.
In any case, we typically have these charts running in the background during the trading day and we are currently working on an Internet-based version to be available to Metal Augmentor subscribers sometime early next year. We believe this will be one of the more powerful market tools available to retail and small institutional investors in the gold and silver markets and so we look forward to announcing new developments very soon.
More for subscribers in the next few days as circumstances dictate.