Thursday, May 15, 2014

High Tech "State of the Union", By the Numbers: Part 5 - The VF Forecast

For any man with half an eye,
What stands before him may espy;
But optics sharp it needs I ween,
To see what is not to be seen. - John Trumbull


Ever since the first cavemen sat around a campfire, grunting at each other as they attempted to figure out where they needed to go for a successful hunt the next morning, humans have attempted to divine the future from a seemingly endless variety of rituals. By attributing hidden meaning to the movements of the moon, planets and stars in the night sky, rolling the bones of small animals, reading the entrails of sacrificed birds or praying for guidance from spirits, people of every era and culture have sought ways to determine what the future might have in store for them so that they could plan and prepare accordingly. 

There is one source of prophetic proclamations which has captured the imagination of historians more than nearly any other. On the slopes of Mount Parnassus in Ancient Greece, there was once a temple dedicated to Apollo, the Olympian diety associated with the Sun, healing, music, poetry and......prophecy. The place came to be known as Delphi, and the Pythia - a woman sometimes of high birth, other times of humble origins - would receive animal sacrifices and alms of gold to deliver visions of the future or advise a course of action on a pressing issue. These prognostications, often obscurely phrased and heavily subject to interpretation, supposedly originating from Apollo himself using the Pythia as a conduit, were eagerly sought by kings, emperors, warriors and philosophers for over twelve centuries.

As an example of the Delphic Oracle's predictions, consider the following: a Greek warrior once asked the Pythia what would happen if he went off to war. Her answer:

"Thou shalt go thou shalt return never in battle shalt thou perish."

Notice how the interpretation of the prophecy drastically changes depending on where one places the necessary commas.

Nevertheless, this sort of forecasting seemed sufficient to satisfy the ancient Greeks, and the Temple of Apollo at Delphi became fabulously wealthy. However, in this day and age, we need something a little more tangible to feel comfortable about our educated guesses regarding the future - data.

Prophecy, however honest, is generally a poor substitute for experience - Benjamin Cardozo

We've seen six full years of quarterly revenue data from "Big Iron" (systems houses), the "Wintel Duopoly", the "Stone Masons" (chipmakers serving the full range of High Tech markets) and the "Area 51" companies of Apple and Google. All of their revenue streams are reproduced below.



(Note: if anyone would like the full dataset in Excel sent to them so that you can run your own analyses, please let me know in the comments.)

There is an inescapable conclusion that one can immediately draw from this table - most of the companies are trending down. This trend is most noticeable with the Big Iron group - IBM, HP and Cisco.

There are three exceptions to this phenomenon:
1. Microsoft - the revenue direction overall seems to be moderately positive. With Microsoft trying so many different things to break out of the declining PC market, perhaps they will find a way to continue this trend. Only time will tell.
2. Qualcomm - the smartphone market has driven Qualcomm's excellent growth these last several years. However, that market is clearly rolling over (please see the March 30 VF blog post "TSMC, Smartphones and the IoT: Behind the Numbers" for a detailed analysis.) Qualcomm's future revenue growth is thus at risk.
3. Apple - the entire mobile computing segment is, to a great extent, a byproduct of Apple's incredible creativity, ingenuity and vision. But as previously discussed, Apple's revenue growth has been gradually plateauing over the last three years. It is likely that this dampening of revenue improvement will become more pronounced. Not only is the smartphone market saturating, but so is the tablet segment (http://news.techeye.net/mobile/tablet-sales-flatten). Both tablets and smartphones are exhibiting classic market saturation symptoms: diminishing returns on feature additions, revenue growth depending increasingly on taking share from competitors as opposed to capturing new customers to the market, and intensifying price wars among smartphone suppliers, as well as between service providers.

The growth of the tablet market is illustrated in the table below. Unit shipments are in millions, on a worldwide basis. (Source: Gartner.)



Notice the year-to-year unit growth rates for tablets: 241%, 93%, 68%. These serial declines suggests that 2014 will be the last year of healthy growth, and that 2015 will likely be more or less flat.

Analysis

It has been the rise of Mobile Computing that has energized High Tech for the last 5 years. The demand for smartphones and tablets cascaded thru the entire technology chain as the need for connectivity and bandwidth drove growth in base stations, wireless backhaul, switches, routers and datacenters, ultimately cascading down the supply chain to chipmakers. The Web 2.0 explosion is itself also largely a phenomenon of mobile computing.

Yet this market - an offspring of the convergence of the communications, consumer and computing sectors - has a critical limitation. Buying a smartphone or tablet is ultimately a discretionary expense, dependent upon disposable consumer income. 

Thus it is macroeconomic factors which impose limits on the growth potential of the mobile computing segment. Examination of those factors leads to an explanation of why this market is now saturating after 5 years of spectacular growth.

Ever since the financial crisis of 2008, a variety of phenomena have combined to dampen a return to general prosperity. Sovereign and commercial bank balance sheets across the industrialized world and the premier developing nations are in poor health, mired in heavy debt loads. The resultant paucity of available credit for business expansion has profoundly hampered the building of any momentum towards economic recovery. As a consequence, unemployment and underemployment statistics are stuck at significantly elevated levels and wages have remained stagnant. If the experience of Japan serves as an example, such a period of tepid economic activity can last for quite a while - in Japan's case, 25 years and counting.

As extraordinary as they are, the novelty of smartphones and tablets have at this point worn off. Consumers with disposable income have made their purchases. In order for them to upgrade to newer models and broader services, or for new customers with less disposable income to enter the market, it is no longer sufficient for vendors to offer enhanced versions of products with more 'bells and whistles.' The vendors of mobile computing products and services will have to offer genuinely compelling value in order to drive their sales back up to earlier growth levels.

That value will have to be quite impressive. It is not only the out of pocket cost that is negatively affecting consumer enthusiasm for new mobile computing offerings. There is also a growing backlash against the technology itself. The following video explains the nature of this backlash perfectly:


Note in particular that this video originates from and is targeted at the most important demographic for mobile computing. 

There are hopes that the Internet of Things will be the instigator of the next growth spurt in High Tech. The trick, of course, will be for consumer electronics vendors to find the right configuration and set of applications that offer a sufficiently enticing value proposition to coax that scarce disposable income out of consumer pockets. The last six months have seen an explosion of devices released to the market from companies large and small, so the experimentation phase is well under way.

Until the IoT can revitalize the mobile computing market, though, High Tech is facing a period of stagnation in all its major markets. Certain segments are in outright decline - note the PC data from the tablet graph above, as well as the data below (in MM's of units) for the worldwide television market (90%+ of which is HDTV.) The market peaked in 2011 and has seen two straight years of decline, with an extended lack of consumer liquidity and continuing problems in the labor market the obvious culprit (source: IHS.)


At least some High Tech leaders are beginning to recognize the problem. Microchip CEO Steve Sanghi sees a period of escalating price competition and M&A activity in the chip business, and is very sanguine about the industry's near term prospects: 
http://www.eetimes.com/document.asp?doc_id=1321761&_mc=RSS_EET_EDT&cid=NL_EDN_DesignIdeas_20140403&elq=a4017030752d4b339b478b2320a8bbee&elqCampaignId=16324

Sanghi is also joined by Zvi Or-Bach of MonolithIC and Henry Samueli of Broadcom in announcing that Moore's Law appears to be failing at this point and there is little if any benefit to be had by going into the deepest submicron process nodes:
http://electroiq.com/blog/2014/03/moores-law-has-stopped-at-28nm/#comment-89083
http://news.techeye.net/chips/moores-law-fails-to-make-cheap-chips
This presents a real Catch-22 for the entire High Tech industry. All the first and second level markets for technology appear to be stalling or plateauing, with a few in outright decline, and the traditional formula of relying on 3P (performance, power and price) improvements at the chip level to drive system level feature innovation for new product offerings is blocked by the effective end of Moore's Law in semiconductor process technology. 

What this really tells us is that the High Tech industry itself needs to change at a very fundamental level. From a long term strategic perspective, things have actually been somewhat simple for High Tech executives for the last forty years - follow the shrinking process geometry curve for significant generational improvements in speed, cost and power, adding features, improvements and enhancements along the way. This product development methodology has long been considered High Tech's standard method of "innovation."

However, it's time for High Tech to abandon this safe historical method of moving the business forward. In fact, the word "innovation" needs to start getting treated like a four letter word. Instead of "innovation", the industry from top to bottom needs to take a cue from its leaders, the Area 51 companies, and adapt themselves to pursuing "invention." 

The word itself denotes something far more significant than improving and optimizing on what you already have. It means creating new and groundbreaking things. These things can be so revolutionary that the world has never seen their like before, or they can be combinations of existing items and capabilities that, in their integrated form, spawn heretofore unseen functions and applications. It is the latter approach that has served Apple and Google so successfully, and it is the former which created wondrous things like the transistor. 

For 2014, the VF forecast for High Tech is that the year will be more or less quiet. My guess is that revenues on the whole will be flat or perhaps 5% down. Beyond this year, though, the going will get significantly tougher. 

For systems houses - particularly those in consumer electronics - the way forward is clear. IoT will take a lot of experimentation to achieve viable results, which is another way of saying that we will certainly see an abundance of product launches that ultimately prove to be failures. Yet progress is impossible unless preceded by many instances of failure and disappointment - the inventions of gunpowder and the electric light bulb are perfect examples of such. Google's Project Ara may play a key role in the development of the IoT. Again, time will tell.

However, truly world shaking, revolutionary change will have to be spearheaded by the foundation, the base if you will, of all High Technology - at the chip level. How that may come about is the subject of a future post.  ;-)


3 comments:

  1. The take away is as Moore's Law reaches it's practical limits, growth due to 3P improvements is not a given. This means semiconductors as a growth engine will be muted. This is not the end of the world as there are solid business models not solely based on growth. As long as semi companies are profitable they can survive just fine, but without growth driving stock prices they will have to increase dividend payouts to shareholders. I'm totally fine with that. Procter and Gamble has been doing it for years.

    ReplyDelete
  2. Actually, Steve, this lack of growth and the death of Moore's Law could create a serious problem. The semi companies will be increasingly forced to compete only on price. Sure, there will be attempts to come up with novel circuit designs to squeeze out a little bit of 3P's improvement for various applications, but those tricks will all get discovered sooner or later, and it will all go back to price. It will drive more consolidation and, unfortunately, bureaucratization of the industry.

    ReplyDelete
    Replies
    1. Yes - agreed, and I think we are kind of there already there. My point is the semis have had their run and the glory days are behind us now. They won't go out of business and will still be profitable, but not growth engines and the darlings of Wall Street. Instead they will simply be another vehicle for paying dividends to their shareholders and I'm fine with that. It's just the natural progression on any industry. But unlike many that have come and gone due to disruptive forces that rendered them completely obsolete and went into a death spiral (e.g., print film, vacuum tubes, vinyl records) I don't see semiconductors just going away. Instead they are just another commodity. Not a particularly exciting place to be, but not roadkill either. This is the world most industries exist in and now the semis are no exception.

      Delete

Feel free to comment or critique!