I made a purchase today that got me thinking about the progress the tech industry has made over the last 30 years -- and, more important, about which developments mattered more than others. The purchase was a home-networked storage device made by Seagate Technology (STX). The device has three terabytes of storage (that is three trillion bytes, with a T) in a small box measuring perhaps six inches tall, perhaps one inch deep. This gargantuan amount of storage cost me all of $179, and was installed in about 20 minutes onto my home network.
Disc-drive-based storage is the unsung hero of the technology revolution. Every major development in the internet economy over the last 10 years was enabled by two factors: huge increases in the speed of networking, and huge decreases in the cost of storage.
The need for broadband download speeds is obvious, but the impact of storage less so. Google (GOOG) could not read, cache, and index every web page in the world without the ability to store exabytes of data inexpensively. Without cheap storage, Amazon (AMZN) or eBay (EBAY) could not store every search, then cross-reference them with other searches and purchases to bring you "people who bought this also bought that." As all media goes online, Netflix (NFLX) and the studios will not be able to store the tens of thousands of movies in their libraries -- at Blu-ray definitions -- without tons of cheap storage. The amount of data stored and cross-referenced is literally beyond comprehension to the human mind -- and is totally reliant on cheap storage.
Most folks are familiar with Moore's Law, which is actually an observation that the "number of transistors that can be placed inexpensively on an integrated circuit doubles approximately every two years." The implication is that the speed of computing also increases by orders of magnitude over the same time frame. While ever-greater computing speeds are critical to progress, they are less important to the development of the Internet economy. If you don't believe so, check your CPU utilization next time you are surfing the web.
The scientists and engineers in the hard-drive industry have been nothing short of brilliant in increasing the "areal density" of drives, which is simply a measure of the number of bits (1 or 0) of data that can be squeezed into a square-inch of disc space. The current state of the art is 541 billion bits per square inch. This compares to the earliest hard drives, which packed in only four million bits per square inch. Storage density, in other words, improved 135,250x over the past 35 or so years. Would only investment managers be able to make the same claim! The very first hard drive was inside an IBM (IBM) RAMAC 305 introduced in 1956, which held 4.4 megabytes in an armoir-sized box -- about the size of one MP3 file today.
The chart below, from FutureTimeLine.net, plots the improvement in storage. It follows a linear exponential curve similar to Moore's Law.
So what is the reward for technological progress that rivals the best of anything ever accomplished by mankind? Not much. The drive industry is notoriously cyclical, often plagued by bouts of oversupply that then give way to somewhat firm pricing -- but never for long, it seems. The business is oddly low-margin, considering how technically complex it is. The major drive vendors rarely command a gross margin above 25%, which is considered a pathetic gross margin in most industries, let alone in technology. They all routinely trade at single-digit multiples, reflective of the low margins and cyclicality. The king of the group, Seagate, was until a couple weeks ago trading at the same price as it did in 2003. A glance at the charts will confirm that these are trading sardines, not eating sardines.
In fact, there are few means to play the group anymore, anyway. There are only five major drive manufacturers: Hitachi Global Storage Technologies, Samsung, Seagate, Toshiba and Western Digital (WDC). Obviously Seagate and Western Digital are the only pure plays. A few years ago, as the industry consolidated, investors thought it would become an oligopoly with rational pricing and better profitability. So far, that has not happened. Perhaps they need to get down to one vendor -- as in CPUs or operating systems!
(By the way, the wildly successful EMC (EMC) -- or other networked storage plays, such as NetApp (NTAP) -- are actually inverse plays. They benefit from the falling prices of hard drives as they stuff tons of them into storage boxes. EMC or NetApp against Seagate or Western Digital is the classic "paired trade.")
One imputed disaster for the drive industry that is yet to happen is the encroachment of solid state "flash drives." A decade ago, the obituary was being prepared for the hard drive industry, as upstarts such as SanDisk (SNDK) made incredible progress on storage density and cost. To their surprise, however, hard drives remained on their own phenomenal density/cost curve. Most likely, flash will catch up eventually, but for now there is no concern about flash vendors putting the hard drive guys out of business. As mentioned, the storage demand is just too high to be met inexpensively by flash.
Investors should continue to view the drive names as trading vehicles, not investments. Having said that, there is precedent for an industry to eventually gain respect among investors. The semiconductor companies used to be considered commodity "parts" companies, and even Intel (INTC) routinely traded at a single digit price-to-earnings ratio as late as the mid-1990s. In fact, there is a famous moment in tech lore when Andy Grove delivered a key note speech at the then-well-attended AEA conference. Grove walked up to the dais wearing a Bill Gates mask, and joked that perhaps that would get him a better multiple!
Of course, Intel now routinely trades at a healthy multiple that reflects its industry dominance. So the Seagates of the world do have some hope, but pricing and margins will need to improve meaningfully and permanently first. Despite that, the engineers and scientists of the industry should be proud of what they have wrought -- and should get rich playing the stocks of the great businesses, like Google or Facebook, that they enable.