click to enlarge
click to enlarge
In mining lore there is a story about miners placing a canary in the mineshaft to test the safety of its atmosphere. If the bird lives, the air's breathable and you're probably going to be ok. If the bird dies, get out quick. The semiconductor equipment industry has their canary in Kulicke and Soffa (KLIC).
KLIC makes the ball bonders, the machines that press the infinitesimal wires of integrated circuits onto their silicon layers. They are a backend semiconductor equipment manufacturer, and 80% of this particular business goes to them. If chips are being produced and the market is strong, KLIC will be adding capacity and manufacturing space to get ready for the orders. They also ramp their orders quickly, so they are a short term barometer on the upticks and downticks of demand for the semiconductor equipment industry.
It's a widely accepted fact that the semiconductor industry is cyclic, and runs in dramatic 2-3 year cycles of boom to bust and back again. Whenever the share price of KLIC has fallen to single digits and closed near its long term trend line (above), the stock has rallied and a new semiconductor cycle has begun. The one exception was the great chip recession of 2001-02; when KLIC broke to ten-year lows (above). The stock has rallied for a minimum of 100% above this trend line, and sometimes for as much as 400%.
The other thing to note is the $5 price line on the second chart (above). KLIC gravitates towards this price in cycle after cycle. In severe downturns, it will go under it and then quickly return to it. In mild retractions, the buy point is somewhere between the ascending trend line (chart 1 above) and the $5 mark. So you buy KLIC somewhere between $5 - $7.50 (where it is now); and a strong buy is anytime the shares are close to $5. Considering the ongoing devaluation of the dollar, and the rising cost of gold (a prime ingredient in ball bonding), it is remarkable that KLIC cleaves to this $5 mark decade after decade.
I first began following KLIC in 1996 (5 cycles ago), and I can honestly affirm what you see on these charts. It's the closest thing to a free or reduced lunch for stock-investing there is. If you begin positions at the right time in the cycle, KLIC can be a very profitable investment indeed.
The economics of the semiconductor industry is best summed up by an insightful observation uttered 30 years ago by Gordon Moore, the past president of Intel. It's called Moore's law, and reads, “The number of transistors that can be inexpensively placed on an integrated circuit increases exponentially, doubling approximately every two years."
Moore's Law comes with a caveat: chip companies must develop their products continuously in order to remain part of this evolutionary development (and to stay ahead of their competition). It's a high-stakes Las Vegas-style poker game, and if you're going to play, you must ante up each cycle. If a company "opts out" or bluffs, they could be through. And the stakes are large: a new semiconductor fab can cost as much as $4 Billion dollars, but that fab could mean the difference between market share dominance or extermination. This is what's behind AMD's ongoing battle with Intel.
One of the most deceptive facets of the semiconductor cycle is the difficulty of assessing our current locus in the cyclic sequence. Right now we have a chip market that's starting to decelerate. Visibility appears cloudy at best. Another fly in the ointment is the overall market - it's near 52 week highs (10.27.07) - and the bottom of the chip cycle always seems to coincide with a bottom in the market. So in my estimation, KLIC closer to $5 may be our optimum buy point. The research firm I have followed for several years thinks the worst is yet to come as the current glut of chips comes on the market in Q4. In July 2007, they reiterated their assessment that a "significant downward turning-point in underlying demand for ICs (would occur) in mid-2007.
The table below shows the sequential percentage returns based on the stock price (including stock splits) beginning from 1979. The percentage changes transpired through the highs and lows of the last 14 semiconductor cycles:
click to enlarge
After looking at the graphs below, you can make the obvious determination that the time to be buying this stock is when NO ONE wants it, and selling KLIC when it reaches a parabola that is at least twice its previous cycle low.
Fig 1. KLIC 1979-92 (weekly)
Fig 2. KLIC 1992-02 (weekly)
Fig 3. KLIC 2002-07 (daily)
- For the last 27 years, investors who have bought this stock at its cyclical bottom have averaged over 200% from the point of their purchase to the peak (+6 to 18 months). It's worth noting that these troughs and peaks have shortened during the last 4 years, and the returns on the upside are now around 90% from the lows.
- Traders who have gone short at the peaks and covered near the bottom have averaged a little less for each cycle, but still have done extremely well.
- Investors who bought at the bottom of each cycle and then sold near the top, have about a 50x bagger. If they did so using margin, they have a 100x bagger or more.
- Traders who went long at the bottom of each cycle and short at the top, and used some kind of margin in the process, would have astronomical returns.
- Investors who bought only once in 1992 and held through all cycles of the decade are up 700%.
So, when is the next best time to buy KLIC with the intention of holding it through one of these cycles? Under "normal" conditions (usually only perceived in hindsight), between now and next April (2008) look good. On a multi-year basis, significant turning points appear to re-occur near the months of October and April. The last time I released this article - summer's end, 2006 - was an excellent time to acquire semiconductor equipment companies. KLIC investors from July, 2006 to July, 2007 received returns along the long-term historical norms illustrated above and were richly rewarded (+92%). But you have to be patient, and willing to begin when no one else is buying KLIC (like now), at a time when visibility is poor. If you've ever watched a tide change along a seashore you'll have a small approximation of what I mean. The shore break becomes disturbed, the clarity of the water muddied. It will take time before the new direction is apparent.
The cycle tops on the graphs (above) are marked in red, the bottoms in blue. For over a quarter century the plot line for KLIC has been the same: from rags to riches to rags to riches; over and over again.
This article was originally written on September 24, 2001. Revised on June 24, 2002; and then updated again on September 9, 2006 and October 27, 2007.