Lessons from 3 days of SLV v GLD

As mentioned in prior posts, I began blogging on this developing Gold and Silver ETF Allocation System via the "CAPS System" on Motley Fool's site, fool.com. Due to the scoring-system used at CAPS though, I am making this blog the primary location online where I will document my progress refining this System and its performance in the market. This entry below was originally posted April 24, 2007 @ CAPS. I've copied it here to provide a complete history to the reader who wishes to research my Allocation System development over time and confirm the transparcency of my process. One can obviously view the original entry from that point in time via my CAPS blog there but I encourage one to primarily follow my Allocation System via this blog, Greham's Day in the Sun.


Last Tuesday, April 17, I posted an early-warning on my blog about what appears to be an upcoming sell signal from my Allocation System. I expected the signal "by or during the first week of May. Within just 10 to 14 trading days from now." That was 5 trading days ago.

For those that haven't read any of my earlier blog posts, this Allocation System I am testing is driven mostly by trends in the Gold/Silver Ratio and also is intended to keep the investor's funds allocated to the tradeable Precious Metals vehicles--like the ETF's GLD (or IAU) and SLV--100% of the time. It is assumed the investor wants this portion of their assets exposed at all times to either Gold or Silver or a combination of both. But they would also obviously like, if possible, to maximize return while minimizing risks within that context of staying fully invested. Therefore, the Allocation System aims to have them more exposed to Silver when the odds favour Silver's outperformance and more to Gold when gold should perform better relative to Silver. Often, when Gold's better relative performance is most notable--and exploitable--is during downturns and market corrections in the metals. In other words, Gold almost always declines less than Silver during corrections in an otherwise bull market for the metals.

Thursday, after my early warning post of two days before, there was a notable one-day correction in both precious metals and in the mining stocks thereof. I've also refered in earlier posts that this System may ultimately be helpful to investors holding precious metals mining stocks to find opportune times to take profits and then re-deploy cash later. Indeed, I think we can generally observe that many of the mining stocks in this sector have generally been modestly correcting already since that early warning one week ago now.

The correction last Thursday was sizeable enough that I wondered if my signal was actually coming too late and that the "early warning" should have actually served as the true-blue Signal. So I went and looked at the historical data. That is what this exercise is all about. Testing and refining. And objectively reporting that process for all to see here.

As mentioned, the Gold/Silver Ratio and relative performance of the two metals is what this system is all about. When a periodic market correction occurs in the metals, Silver almost always declines more than Gold. I knew that during the most recent periodic cycles (dated in earlier blog posts), when the correction got into full swing, that Silver sold-off sharply and very quickly. To avoid this very steep decline in Silver by taking relative safe-haven in Gold is the most important goal we must accomplish with this system to outperform a buy-'n-hold strategy using either or both of the ETF's. Last Thursday, SLV's decline was 1.863 times that of GLD's. That seemed very normal and I also suspected it was not the kind of behaviour we've seen in recent corrections. Sure enough, during the prior two intermediate corrections in this market, the one day sell-off in both GLD and SLV that serves as the demarcation point for us, SLV's decline was 4.169 times that of GLD's during one cycle and 5.135 times during the next cycle. At least these hisotrical markers indicate that when the larger expected periodic sell-offs come, we would expect one of the major characteristics to be a much larger relative decline in SLV than what we saw last Thursday.

It appeared from my research that we should not probably have expected a follow-through and increased selling of SLV and GLD last Friday. At least if my System is generating signals accurately. And sure enough, both ETF's corrected back from Thursday's decline and made-up some of their losses from Thursday in Friday's trading. GLD rose 1.7% and SLV 2.3%, true to its character and historical performance. When I look back at recent similar-appearing stages in the cycle, this volatility is not uncommon. In fact it appears to be more the norm. Preliminarily, after looking at more data this weekend, I knew that what we should really suspect is that sometime before the 2nd and major Sell Signal is generated moving us completely out of Silver, that we should expect SLV to outperform GLD. Trading Monday April 23, came-in right on que it seemed. Monday, while GLD actually declined on the day, SLV posted a gain.

Now, I must admit that possibly the biggest challenge for me refining this system is the timing of the calls and the "last" or "2nd" call in particular. Frankly, I'm nervous as heck right now that the "real" correction could begin any day now. I will be quite surprised if we get through next week, the first week of May, without the correction getting into full-swing. So partially, I want to preserve gains and error on the side of being "early". But I'm not sure that is really what the System is saying to me so I'm resisting the urge, for now. Of course, this is "all off" if some exogenous event comes along on the major economic or political stage that disrupts what we might otherwise expect to be "normal market conditions" in the metals right now. By definition, I don't expect any such disruption though. Precisely because it would be un-predictable event.

Every sign is though, to me at least, that we are in the waning days of this short-term uptrend that began on or about March 1st this year after the last minor correction in the metals. For instance, Silver was much higher in Monday's session and came back down to close with a much more modest gain that it showed earlier in the trading session. Watching the action of the metals mining stocks these past days also seems to indicate some "toppy" signs. For instance, two Silver Stocks favoured by many investors in this sector showed weakness today (SSRI, PAAS). The ETF GDX, which attempts to match the AMEX Precious Metals Miners Index, declined almost 1%

With SLV giving back much of its gain Monday and GLD showing an actual decline, this market is looking overdue for a correction.

In subsequent posts I need to report on several items:

1. Adjustments that I now see I need to make to better match this System to the scoring rules used in Motley Fool's CAPS;

2. Documenting a couple of trades I've most recently made in light of #1 above;

3. Observations I'm making that the relative out-performance of Silver has been weakening each of these past cycles in this market since late last Summer (2006). That in fact, during this current cycle, the Gold/Silver Ratio has tested it's 200 DMA several times and appears, upon any decent correction, to be positioned to pierce the 200 DMA and that too, may indicate a coming period where Gold and Silver perform more similar to one another and, in fact, Gold could even outperform Silver for awhile.

4. The beginning of my research into the operational stability and merit of the Precious Metals ETF's and pros and cons of these market instruments compared to the investor actually taking physical possession of bullion. And finally, news on more physical bullion-backed ETF's coming to market and this time, the long rumoured ETF's for Platinum and Palladium. I will also attempt to include my thoughts on whether the continued creation of such ETF's could be a contrary indicator about this market.