Calling the top on anything is rough stuff. We have seen top calling in so many places at so many times in the last few years, that it becomes commonplace, noteworthy even if it is wrong. One look at Tuesday's robust housing number, as well as the recent auto figures, tell you that the long-called peak housing and auto theses have not played out.
Sure, you could argue the auto stocks haven't made you much money since the top-calling began in earnest. But General Motors (GM) has stirred, and the original equipment parts suppliers have seen their stocks roar.
We have endless top-calling in Amazon (AMZN) -- all for naught, given how the company's stock seems to be headed to $2000. Tesla's (TSLA) stock never collapsed as so many top callers demand to happen. Failure to comply with the conventions of financial metrics, Elon Musk style.
And then there's DRAMs and Flash memory prices, the subject of which caused some very real consternation among complacent owners after the highly accurate, very rigorous Katy Huberty, the doyenne of tech, on Monday called a top in flash pricing, rocking the group. Shareholders in stocks like Micron (MU) , Western Digital (WDC) and Seagate (STX) , Applied Materials (AMAT) and Lam Research (LRCX) among them, must all understand the dynamic that's now playing out before you pull the trigger.
Flash and Drams are two commodities -- although the makers of both of them would beg to differ --that frequently give you boom-bust moves of breathtaking proportions. The upside comes from a surge in demand without a concomitant increase in stock price, as canny investors suspect there is too much capacity coming on. Then, when demand exceeds supply long enough, the producers all gear up for new lines to grab the excess margin.
Most Wall Street analysts believe these competitors can put plants up fast enough to capture that marginal dollar without a peak developing. But they are always wary that we might have reached that tipping level. The companies themselves have been much more worried about trying to meet skyrocketing demand for every trickier-to-build device than they are about restraint on expanding their own product lines. That's what happens when a tidal wave of new devices, which tend to need storage, short and long term, cloud or no cloud hits a very small industry.
It's always hard to gauge when or even if that demand top is upon us, because near the top the clients all hedge their positions and take down excess inventory. This is the phenomenon known as double ordering, and it continues until you actually get what you want, and shortly before that prices tend to start going down, but they accelerate when the double orders evaporate.
Against the double-ordering threat is the lengthening cycle of building new supply, in part because, as I mentioned, these semiconductor equipment companies aren't stamping out auto parts. They are creating highly sophisticated devices that would never have been thought possible even a few years ago, as their complexities and size limitations make them incredibly difficult to build.
There have been many top callers in this cycle, in part because it's now been a couple of years of halcyon pricing, the usual half-life of these moves, and because the spending to build more machines is off the charts. It's just presumed that either situation will create a virtual crash of pricing, and that's why the stocks of Micron and Western Digital sell at such low price-to-earnings ratios, namely, around six times earnings for both.
Only the stock of General Motors rivals those feeble numbers. That's a sure sign that while Wall Street is exceedingly bullish, the investors and traders don't trust next year's numbers. Judging by the gap, it's almost as if they think earnings will be cut in half.
Now, in the most recent quarter DRAM pricing continued to accelerate without pause, despite a continual but disciplined amount of equipment spend. Not so with flash, which saw a slower rate of change of growth, dreaded because that bit of the second derivative, barring some glitch of complicated manufacturing, assures us of a top down the road.
Flash capacity spending is far exceeding DRAM spend this year, because of a perception that flash remains the storage of the future, even as both kinds of chips can play the role. But no one in the real world, where customers are worried about not meeting agreements, thinks about the stocks. They think about the prices and watch their profit margins shrink.
Management of all the companies in the supply chain are all on tenterhooks to see if those trends continue, as their revenues depend on the same calculus.
But here's the key distinguishing point: Micron has less on the line than Western Digital right now. That's because for most of the last decade, Micron has spent aggressively on building out and buying flash memory equipment makers to lessen dependence upon the historically more voluble DRAM component.
Western Digital hails from a disk drive world, which suffers from its own notorious boom-bust cycle; so wisely, last year it spent $16 billion on Sandisk, a company with a key partnership with flash maker Toshiba.
By this point, you have heard about the wrangling by Western Digital to buy the other half of its joint venture with Toshiba, in part because Toshiba's in such trouble related to nuclear power plant construction, for fire-sale prices. Toshiba doesn't recognize WDC's ownership, claiming its deal with Sandisk was relinquished if Sandisk were acquired.
The fight seems to have died down now, even with Toshiba making a deal with a different entry than Western Digital -- something WDC tried mightily to prevent from happening and still questions its legitimacy.
With Monday's gigantic selloff, the market's reflecting that Western Digital will now be more of a hard drive maker than investors thought, which means a lower-price-to-earnings multiple, given that flash had the highest place on the hierarchy.
If we get a real peak, Huberty writes, you could see a dramatic decline in pricing and a tag-a-long crushing of Western Digital's stock, but not Micron's because, ironically, DRAM pricing is still firming. It's ironic, because it was always thought to be DRAMs that collapsed first, not flash, because the former is more quickly and easily made, so capacity is easier to add.
Is the call right? I think it's a judgment based on a holistic approach matched with empirical evidence which yields a belief that the upside will be constrained anyway and the downside dangerous. I think that's a decent calculus.
Several months ago, at around the price shares of Western Digital traded coming into Monday's session, we called a top for Action Alerts PLUS club members as part of a rare trading strategy when we try to teach longer-term, research-based investing. Perhaps it is telling that as the price increases continued in the real world for WDC's wares, its stock price stalled out and didn't go higher -- although, had it got Toshiba's portion of the Sandisk partnership, it most certainly would.
I do believe that this cycle is different, though, because of the speed with which the data center must be built out to meet demand coming not just from e-commerce, but every single industry's internet of things trail. In that sense, I think the growth is more secular in nature, something that Lam Research CEO Martin Anstice and Applied Materials CEO Gary Dickerson have emphasized when they came on Mad Money.
In some ways, their theories are hard to argue with. If the amount of data is exploding, so will the devices needed to service them. That's what makes me suspicious of the peak call. How can you think that this time isn't different, given the step function thinking embraced by pretty much everybody?
Would that mean I would buy Western Digital if it fell back to the low $70s, where we first started to accumulate it? Yes, if pricing it stable; no, if it's going lower.
Either way, just remember that my thesis, the secular growth thesis, is really only embraced by the people who have the most at stake, the equipment producers, even as I think they know more and will be right. The group's stocks may no longer go up with great ferocity, but then again, when they are down, they may be must-buys.
