Tech Stock Warning
The following is an excerpt from Marathon London Global Investment Review volume 13 No. 8 December 31, 1999, (USA section).
The aggregate market capitalisation of the forty-five or so shares in the Marathon US portfolio is close to U$800bn, and earnings for these companies this year will be around U$54bn. The market capitalisation of Cisco, AOL, Sun Microsystems, Yahoo! And Amazon.com, (a sample from the internet economy) is also around U$800bn and earnings this year will be U$4bn. The Marathon portfolio has a p/e ratio around 15x, the Internet basket around 200x. Stock market performance of course has been found (almost) entirely in the latter group in 1999. We have argued the case for investment in Marathon’s portfolio exhaustively in the past (declining capital spending, shrinking share count, industry consolidation, low starting valuations, rising demand etc.), but have not specifically argued the case against investment in technology. In the August 1995 edition of the GIR (volume 9, number 5) at the height of the Asian bubble, we presented eighteen reasons to be bearish on South East Asia. Whether or not this quarter is the peak in technology shares, (the bad news is we were eighteen months early on South East Asia), may we now present Marathon’s nineteen reasons to be bearish on technology in December 1999:1. Valuation; technology stocks have risen since the mid 1990s to discount earnings growth of 14% over the next ten years, according to Sanford C. Bernstein, an investment bank. This is up from a discounted earnings growth rate of 8% three years ago, and represents a higher level of expectations than the 1983 peak, and compares with just 4.2% for the market at large. The same conclusion is reached on the other metrics: at the beginning of the decade Intel sold at 2 times sales. The stock recently peaked at 10 times.
2. Valuations are also high relative to the consumer’s ability to pay. In the last edition of the GIR (volume 13, number 7) we argued that household expenditure on the new digital economy would need to rise from 21% to 35% of discretionary income by 2001/2002 to justify current valuations. The effect of course is that the money would need to come from elsewhere. Perhaps people will drink less beer, take fewer holidays, and generally change their lifestyles to accommodate their appetite for digital services (mobile phones, the Internet, PAY-TV), but we doubt it.
3. Technology firms enjoy high return on assets and equity as recorded in company accounts but it is worth noting that these returns are not adjusted for technology’s short product life cycles. For example, a software product which lasts for three years and earns 40% on invested capital, fails to break even in economic terms, if one assumes a cost of capital of 10% (actual return: negative 0.5%pa). In other words technology profitability is high, but it is not as high as it looks.
4. Investors in Internet firms may wish to note that the historic success rate of investors in nascent industries is not encouraging. Warren Buffett pointed out in a recent Fortune article (November 22nd, 1999) that whilst the aviation and auto industry changed society, this was of little benefit to investors in these industries.
5. The effect of high valuations is to encourage high levels of reinvestment and rising levels of competition, which may lead to declining industry returns. Barriers to entry in many technology businesses are modest in terms of invested capita, and are largely limited to patents, licenses and access to capital. The latter is not a problem when the implied cost of capital for many internet start ups is close to nil, and according to Wells Fargo 25 internet start ups are being funded a day in Silicon Valley.
6. The problem with patents is that the patent infringement litigation process may be longer than many product life cycles and so good ideas risk becoming dissipated amongst the competition.
7. One response by the incumbents to rising levels of competition has been to acquire the embryonic competitors. This is the approach followed by Intel, Microsoft and Cisco. However, the incumbents have built their past success and high levels of profitability on organic growth, supported by in house R&D and capital expenditure. By migrating to an acquisition premia paid. The actions of the incumbents (the purchase of promising new technology firms) also helps promote speculation in the IPO markets.
8. High share turnover. The average holding period for many technology shares is one calendar quarter. If the outlook for technology has never been better, why do investors own the shares for such short periods? We draw the parallel between the short holding periods in technology in the US today, and that of the Malaysian Second Section (the speculative end of the Malaysian stock market) in the period immediately preceding the Asian stock market crash of 1997. Here too holding periods were as short as one quarter and secretaries also resigned to pursue careers as day traders!
9. The compression of investment time horizons employed by many Wall Street analysts. This applies to all industries, and leads analysts to recommend shares in firms where the current business (correction: earnings) trends are already favorable. Whilst this trait has always existed to some extent, it seems that investment time horizons have tended to shorten with the duration of the bull market. Stock recommendations have therefore tended to reinforce the prevailing share price momentum regardless of price (and so presently favoring technology shares), rather than highlight longer-term value (elsewhere). In a recent discussion with a US insurance analyst, Marathon and the analyst agreed that Aetna, a company whose share price has halved this year and which intends to repurchase 15% of its shares outstanding by the end of 2000, has a normalized value in excess of twice the current share price. Should such a price target take five years to achieve then the annual return would be around 15% from current levels, not too bad one may conclude, especially compared to the prospective return from the Dow over the next five years. The analyst’s recommendation: sell, on the grounds that earnings growth would likely be muted in the next two quarters. "You could have made 40% this week in Yahoo!" she added.
10. The all pervasiveness of the new gold rush. This relates to the blanket coverage of technology issues in both the business and now the mainstream press, almost entirely favorable and ebullient in their blue-sky scenarios. Perhaps void of visions for the new Millennium, media editors are using technology to fill the gap. It is interesting to note the endorsement by Harvard professors of Internet start up companies formed by their students. When academics are seeking to monetise their standing in the stock market the top is unlikely to be far away. We ask, would they have also endorsed a student with a plan to improve the Bessemer process?
11. The increased public appetite for speculation. In July 1999 Barberis, Huang and Santos published a paper (National Bureau of Economic Research, working paper 7220) in which they argued investors become emboldened by prior stock market gains and take greater risks with the profits ("House money") than they would with their original capital. We have much sympathy with the findings, but would note that the process cannot endure indefinitely: Risk is not the only factor necessary for wealth creation. The technology sector receives the bulk of the reinvestment of house money, as any cursory visit to the stock chat rooms on the Internet will testify. For evidence of the power of this money to move the markets look no further than recent trading in Analogue Devices. The company reported better than expected earnings on December 1st and the shares rose U$2 and U$1 the day after. At 2:58pm on the second day an item appeared on the Yahoo! Investor’s bulletin board reporting the favorable quarterly result and the stock immediately rose a further U$5 (8%).
12. Equity issuance is high, not of itself bearish, but indicative of high valuations, and of course is in contrast to the declining share count to be found elsewhere in the market. Of the 212 U.S. IPO prospecti received by this office in 1999 (a shot in the dark if ever there was one), 60% were from technology and communications firms.
13. Dilution from the future equity issuance. Firms, which lack cash flow, are forced to pursue capital market dependent business models. Take for example Williams Communications which is laying a fiber optic network. The company knows what capital spending will be this year and next as the funding from the equity and bond markets has already been secured. But what of capital investment in 2001 and 2002? Without the funding management cannot be sure, at least with any accuracy. At the current valuation (approximately four times invested capital) the dilution from future capital raising may be modest. The problem comes should the capital markets be less generous which raises the prospect that dilution could be considerable if the network is to be completed. The risk, not reflected in current share prices we suspect, is that the level of dilution is unknown.
14. The relative euphoria of equity investors compared to bond investors. Businesses such as Williams Communications are valued by the equity market at multiples of invested capital whilst their bonds are rated at junk status. These two views are not necessarily contradictory but imply equity investors have made a quite different, more generous, assessment of the likely cash flow from the business, compared to their bond market counterparts.
15. High levels of institutional ownership. Growth funds have tended to be overweight the sector, partly as growth has been hard to find elsewhere in the market, and increasingly traditional value investors have followed in their wake. Positions may also be entrenched: one large, Boston based, mutual fund manager has over 40% of a US portfolio in technology shares alone and recently point blank refused to entertain the thought that the shares may be expensive when challenged by a respected Wall Street analyst. Their defense was that they had not spent two years building these positions to change now!
16. The narrowness of earnings growth in the market will pass. As we reported in the last edition of GIR (volume 13, number 7) one reason technology has been so strong has been the lack of earnings growth elsewhere in the market. This process is likely to prove cyclical, rather than secular, and implies that the valuation distortions in the market are also cyclical. The effect however, has been to create an almost Zen like symmetry to stock market valuations, as for every overvalued technology share there is a modestly valued, old era business (the money for technology share purchases had to come from somewhere). When shares in companies of the quality of Disney, Xerox, Coca Cola, Gillette, Mattel, Kroger, Albertson’s, Progressive Insurance, Lockheed Martin, Compaq, and Waste Management have declined by up to two thirds, the investment opportunity in our opinion no longer lies in technology shares.
17. The rise in technology shares has been due to changes in price (price: earnings ratio expansion), more than earnings growth.
18. According to Bear Stearns, an investment bank, the ratio of inside sellers to buyers in technology is at a record high. This is in contrast to US firms in general where director purchases currently outnumber sales by two to one. One could also point to the examples of Mr. Steven Case (Chairman and CEO of AOL) and Mr. Bernie Ebbers (President and CEO of MCI Worldcom) who purchased large positions in Maui Land and Pineapple and Joshua Timberlands respectively this year. Neither of the purchases are of new era businesses.
19. Technology firms are arbitraging their high valuations to more modestly valued firms, especially if the acquisition target has customers. For example the contested acquisition of US West (an RBOC with customers) by both Frontier and Qwest (wholesale optical fibre networks), the purchase of TIR, a global institutional stockbroker, by E-Trade (on line trading), and Sound View Financial (a bricks and mortar stockbrokerage and investment bank) by Wit Capital (an on-line investment bank). Of course all purchases have been paid for in stock.
This is not to argue the imminent collapse of technology share prices. Unlike Asian shares in 1997, technology businesses are generally robustly profitable (for the duration of this product cycle at least) and balance sheets are largely ungeared. Demand of course is growing too. But none of this, in our opinion, necessarily makes the shares good investments at current prices. Indeed some of the best e-commerce investments may well turn out to be established franchises trading at 15x earnings in the Marathon portfolio rather than the pure plays on 200x earnings in the NASDAQ. Take for example the rating agencies (how else do you know who you are dealing with over the Internet?) such as Moody’s owned by Dun and Bradstreet) and electronic transaction processors (such as First Data) both of which are included in Marathon portfolios. For some, the new Millennium also ushers in a new era, but one whose principal characteristic it seems, is a low cost of capital for its proponents. We recall similar cases being made for the Asian miracle in the mid 1990s, but of course, investors in the US today are much more rational. Aren’t they?