Giving Up on Semi Stocks

Bottom Line: Two unique measures of sentiment towards the semiconductor group suggest further losses should be relatively contained.

One of the indicators that has been a good guide this year is the Liquidity Premium we post for the SPY and QQQ exchange-traded funds, particularly so the one for SPY. Recall that the theory behind the Liquidity Premium is that as traders become more and more uncertain as to future market direction, they will reduce their trading in individual equities and instead concentrate on the ETFs. With no short-sale restrictions and extreme liquidity, they are an excellent alternative to stocks. When we compare the volume in the ETFs to the underlying equities, and we see that the ETFs are getting more than their fair share of volume, then we have a pretty good idea that there is a healthy dose of uncertainty in the air.

I think this concept is sound, and if it is, then we should see similar effectiveness when we look at other ETFs and the volume in their component shares. I have looked at a couple of other popular exchange-traded funds and the most promising is that for semiconductor stocks.

The chart below shows the Liquidity Premium for the semiconductor group, SMH, over the past year. The indicator is computed by comparing the volume in SMH to the volume in the 20 underlying stocks, and will rise when traders are favoring SMH over the individual equities, and will fall when the equities are getting the bulk of their attention.

SMH Performance After Premium Extremes 09/03 – 09/04
	When the Premium is Greater Than 25%		When the Premium is Less Than 0%
	10 Days Later	20 Days Later	10 Days Later	20 Days Later
Average Return	2.5%	3.8%	-2.0%	-6.3%
% Positive	62%	74%	34%	10%

From the table, we can see that there was a wide discrepancy between future SMH performance after times when the Premium was high versus when it was low. Looking out 20 days, the index was positive nearly three-quarters of the time, sporting an average return of almost 4%, when the Premium was 25% or above (there were 34 days total). On the other hand, looking at future performance after the Premium went negative (68 days total), SMH was positive only 10% of the time and the average return was a dismal -6.3%.

I have only included the past year of data for a reason. Prior to that, the relationship between the Premium and future performance was not nearly as clear, and I believe the reason is volume. The main assumption behind the concept of the Premium is that the exchange-traded fund has to be liquid enough in order to be considered a viable alternative for traders of semiconductor stocks. Prior to September 2003, the average daily volume of SMH was just over 4 million shares. However, from September 2003 through today, the average daily volume has lept to nearly 17 million shares – a four-fold increase. Looked at another way, volume in SMH equated to about 5% of the total volume in the underlying stocks prior to last year. Now, however, SMH volume regularly exceeds 10% or more of the volume of its underlying stocks. This is a vital distinction and most likely the reason why the Premium became more effective over the past year.

Currently, the Premium for SMH is mixed. There has been tremendous volume in the fund over the past couple of weeks, setting all-time records, but volume has also been relatively heavy in the underlying shares. When compared to each other, we do see traders favoring the ETF, but not quite to the extent they did at some of the other lows over the past year. If we could see a few days of extremely high ETF volume (in relation to the volume in the underlying components), it would suggest that we have seen enough uncertainty regarding the group that further downside should be minimal.

That would certainly coincide well with what we see in the Rydex Electronics Fund, the Rydex mutual fund that closely tracks semiconductor stocks. The chart below shows the total assets in the fund over the past year.

Stressing the skepticism surrounding the group, assets in the fund barely budged when semi stocks exploded for a 13% gain in mid-September. As a matter of fact, assets in the fund stood at $36 million at the low on September 8^th. After the rally, on September 21^st, assets were at $32 million, a decline of $4 million after a 10%+ rally in the shares. Assets have been this low only two other times in the fund’s history – October 2002 and February 2003, both good times to take a contrary stance on the group. Such behavior is an excellent sign that apathy towards the group is becoming excessive.

Semi stocks are an interesting barometer of investor sentiment. We usually read either about traders loving them or hating them, and there never seems to be much room for middle ground. From these measures, it looks like the middle ground is probably the place to be - over the intermediate-term, we shouldn’t see a prolonged decline in price, but we’re also not seeing the outright hated that normally accompanies a good time to buy. We’ll continue to watch our measures and see if perhaps they can give us a clue as to when that buying opportunity may be closer at hand.

Since 1997, the S&P 500 has rallied 0.5% or more two days before quarter-end 12 times. The following day was higher 8 times, with an average return of 0.0%. However, the day after that (the last day of the quarter) was higher only 3 times, with an average return of -0.3%. The most positive day ended up being the 2^nd trading day of the new month, which was higher 10 out of 12 times with an average return of just under 1%. I’m not sure how much weight to give those statistics, but it does bear out something I have said repeatedly – if you’re counting on “window-dressing” to push prices higher, it’s best to rely on something else. There are probably some individual stocks that show evidence of quarter-end markup games, but the broader market does not.

Last week I showed that it was rare for the type of dramatic reversal we saw to lead to anything severe on the downside. If the broader market can continue to recover in the coming days, when combined with the skeptical sentiment we’ve seen towards the rally all along, I believe it bodes well for a continued rally in the weeks ahead. Even if we don’t keep rallying from here, a break below last week’s lows will most likely lead to a fitful bottoming process, and we should not violate the August lows.

This disclosure is not intended as trading advice in any form. It is meant as a note to subscribers that the author may have a position directly affected by the market outlook reflected in the commentary. Although the author takes great pains to remain objective in any commentaries, it is only fair that readers should know that the author may have taken positions in accordance with his market outlook. Positions can and do change at any time, without notice to the reader.