Speculators Alter Their Positioning During the Post-Crash Week
It was slightly mysterious why speculators apparently had only liquidated a little bit of their positions on a net basis in the week gold suffered its "mini crash." Some people have suggested that there was perhaps a problem with collating the data in a timely fashion, as the positions were evidently altered quite a bit in the following week. On Friday, the CFTC published the new commitments of traders report showing the situation as of Tuesday last week.
On the surface, it appears as though speculators were liquidating, respectively re-shorting gold and silver in a week when prices rose quite strongly from the crash lows made overnight on Tuesday the week before. It is actually unfortunate that we cannot be 100% certain that this is what happened, as it seems at least possible that the above mentioned data collection problems may be behind this curious development (we have no indication that such problems actually occurred – it is an idea that was mentioned inter alia by Steve Saville this week and it seems to us that it may have merit based on our experience with the legacy CoT data and how they normally behave).
It is unfortunate because it would have been quite normal for this to happen in the week of the crash, but it is rather unusual for it to happen when prices rise. Be that as it may, since there has been an unusually large change in the data, we want to show these changes and offer our interpretation.
First, let us take a look at gold itself. Following is a 30-minute chart showing the action over the past week, and then a daily chart that shows where and when exactly the post crash low was made.
(click to enlarge)Gold, 30 minute chart. The latest CoT report's cut-off date is Tuesday April 23 – so the changes in the report are supposed to have occurred between the day the intraday low was made (April 16) and April 23. Gold rose by nearly $100 during this time.
The daily chart of the June 2013 gold futures contract – a low just below the level of $1,330 per ounce was established in overnight GLOBEX trading on Tuesday April 16.
Before we get to the CoT report data, we want to quickly mention why we concentrate on the aggregated futures only legacy report. There is for one thing the matter of continuity – since we are used to the legacy report, it is easier for us to interpret it. The main difference between the legacy and the disaggregated report is that the latter breaks out the actions of swap dealers and producers/merchants (which together make up the "commercial hedger" category), as well as "managed money" and "other" big speculators in the big speculator category. Note that to some degree, these classifications overlap. A merchant can, e.g., also engage in swaps or enter a speculative position.
It is certainly useful to look at the disaggregated report as well, but generally speaking, there is no big difference in terms of the signals given by the two types of reports. The reason why we exclude options from the analysis is that one can engage in four different types of positions in options: long or short puts and long or short calls. If one only knows that someone is "short" or "long" options, one does not actually know whether they are positioned for a rally or a decline. Someone short puts is, e.g., bullishly positioned, while someone short calls is bearishly positioned. Since these details are not made clear in the combined futures and options report, we rather concentrate on futures only.
Let us now look at the data as well as the historical CoT charts of the futures only legacy reports of COMEX gold and silver futures. Below is a table of the gold CoT report of last week, with a few brief comments.
The gold CoT report of last week. The most important points: big speculators liquidated some of their gross longs and added massively to their gross short position; small traders meanwhile ended up flat for the first time since early 2001. These actions resulted in a near 38,000 contracts swing in the net commercial hedger position (the net short position of producers and merchants alone – this is to say, commercial hedgers less swap dealers – is now a mere 50,000 contracts).
A chart of the legacy report positions shows that small traders have utterly capitulated.
In order to show the larger context, here is a long-term chart of the commitments of traders in COMEX gold futures. Small traders haven't been this bearish since 2000/2001.
The net position of small traders is the lowest in almost 13 years.
Very similar developments occurred in silver futures. In fact, we believe that the net long position of small traders in silver futures may have reached a record low. We only have the data back to 1998, but since then, their net position has never been lower than it was last week. Since 1998 was also close to the time when silver made its secular post 1980 low, it is actually quite reasonable to suspect that a new record low in the small trader net position was set last week.
This is quite stunning, as small traders are traditionally the by far most stubbornly bullish bunch in the silver market. However, the recent move lower has clearly shaken them – more than the crash in silver after the April 2011 high or the secondary crash of November 2011, both of which actually entailed much larger moves in the silver price in both absolute and relative terms.
Last week's commitments of traders in silver – the net long position of small traders has probably reached a record low.
Commitments of traders in silver: small traders have never been as badly shaken up as they were last week.
What to Conclude
In combination with recent sentiment surveys, which show traders as well as gold newsletter advisers at their most bearish since the late 1990s leg of the preceding secular bear market, the new positioning data should probably be interpreted as bullish overall – albeit with an important caveat. As is usually the case with technically and market structure based analysis, it is unfortunately not possible to arrive at an unambiguous conclusion from these data alone.
For one thing, money managers – and the big speculator category generally – are not exactly dummies. Many base their decisions on the market's technical conditions and it is clear that gold and silver both have broken very important levels of support. Therefore all systematic traders and trend followers (i.e., traders guided by "black box" trading programs) are currently probably short, with the exception of those who bet on "reversal to the mean" trades (there are a few trading programs that are concentrating on such reversal trades, this is to say, they go long in order to capture short-term bounces after crash-like moves in stocks or commodities).
It is therefore concerning that this group has increased its gross short positions. On the other hand, the four largest traders that are registered with the CFTC are now holding more than 28% of the gross long positions versus just above 24% of the gross short positions of the total open interest in gold futures. That is rather unusual as well.
One must also keep in mind that the underlying stock of "strong hands" type speculative traders remains engaged on the long side in gold. After all, the net speculative position of the big trader category remains long to the tune of 100,000 contracts (we estimate that the long-term investors that have remained net long through all upheavals in the gold bull market to date hold about 80,000 contracts). These traders apparently continually roll over their long positions, regardless of short-term gyrations in the gold market. We would become very concerned if these traders were to change their opinion, but it is important to note that they have not yet done so.
Small traders meanwhile have now adopted a net "flat" position – neither long nor short – for the first time since early 2001. In the course of the gold bull market to date, it has been a pretty reliable rule of thumb that whenever the ratio of shorts to longs held by this category of traders went to about 2:1 in corrections, it was a good time to buy. The sole exception to this rule was previously seen in the large 2008 correction (which was incidentally of a similar size in percentage terms as the recent one). However, even on that occasion, their position remained net long, even if the size of the net position had shrunk considerably and the ratio of longs to shorts went below the 2:1 level. A flat position is definitely a new phenomenon. So, is this an unreservedly bullish signal? We would answer this question as follows: if the secular bull market in gold remains intact, then the answer is yes.
However, if either a secular change in trend has occurred or a major cyclical bear market is in train (similar to the 1975-1976 mid cycle bear market in the 1970s secular bull), then the small speculator position can be expected to either remain flat or even move to a net short position while prices continue to decline.
Prices and Fundamentals in Secular Bull Markets
Finally, we want to make a brief comment on how secular bull markets in financial assets usually end. Note here that this does not amount to a "hard and fast rule" that is grounded in sound theory. Rather, it is an observation on market psychology based on empirical data.
If we look back at major secular bull markets of the past, such as the Nikkei bull market from 1969 to 1989, the Nasdaq bull market from 1974 to 2000 or the gold bull market from 1968 to 1980, a few things could be observed in all of them as they entered their final stage.
In all these cases, a vast expansion of the money supply and too low administered interest rates played a major role in the formation of the bubble stage. Also, in all cases the biggest gains were achieved in a final "parabolic" advance, which exhibited annualized rates of change ranging from 80% up to 200% (the latter incidentally happened in the case of gold's final ascent in 1979/1980). In short, the by far biggest gains were made in all these bull markets in a very compressed time period shortly before they ended – during what one might term the "bubble' or "mania" stage.
However, two other important aspects unite all of these bull markets: first, in all cases, a major correction occurred just before the bubble stage began. In the Nikkei's case, it was the 1987 crash. In the Nasdaq's case, it was the 1998 crash during the Russian/LTCM crisis. In gold's case, it was the 1975-76 cyclical bear market.
The other aspect is that just as the final, and most stunning price increases were recorded, the fundamental backdrop had already clearly begun to deteriorate.
Contrary to widely accepted lore, not one of these markets proved capable of "discounting" deteriorating fundamentals in advance – on the contrary, they produced their biggest gains well past their fundamental sell-by date, i.e., their final advance lagged the fundamentals (the same by the way happened in the stock market mania of the 1920s – by the time the DJIA made its all time high in September of 1929, the U.S. economy was already in recession).
In all these cases the major "fundamental" datum that changed for the worse were nominal and real interest rates. Investors and traders pumping up the Nikkei in 1989 thought they could ignore the fact that the BoJ was belatedly hiking rates and that JGB yields were rising strongly. Gold traders didn't believe Paul Volcker would succeed in cracking the inflationary psychology of the 1970s by raising interest rates and stopping the expansion of the money supply. Nasdaq traders believed that the magic of the Internet had made technology stocks impervious to a tightening of monetary policy by the Greenspan Fed.
We mention this because it could become an important feature of how the current secular gold bull market ultimately plays out. If it plays out in similar fashion, then it is clear that it cannot have ended in November 2011. On the contrary, this would mean that the by far biggest price gains have yet to happen.
We personally believe that to be the case – but we must, of course, warn readers that we may well turn out to be wrong about this. There are no guarantees – all we have on our side is history and what amounts to an educated guess. Oh, and Ben Bernanke, of course.
Charts by Sentimentrader, BarCharts; Tables by CFTC