The Value Investor’s Technology Dilemma

Published on October 23, 2009

San Jose, California, September 24, 1998. Just another routine day at the office in a suburban office park in the Silicon Valley was interrupted by loud sounds of a celebration coming from the common hallway serving several companies in the building.  Was this a birthday party getting out of control, or some other disturbance?  Actually, it was the day of eBay’s initial public offering.  At the time, I was working as a software engineer for another company operating in the same office park and witnessed some of the festivities.  Based on the initial spike in the stock price, it was easy to see why some of the early employees were celebrating.

Due to the close proximity of eBay’s office, online auctions were often discussed around my office.  I was very familiar with the concept, the internet, and the company.  I would like to say that I failed to participate in the IPO due to application of rigorous value investment standards, but this would be revisionist history.  Instead, I was convinced that the idea of online auctions was crazy and could not understand why anyone would want to rummage around for garage sale items on the internet. I simply dismissed the business as a fad. My co-workers and I made fun of the crazy auction guys down the hall.  I never even read eBay’s financial statements, which I now regret even though it would not have changed my decision not to participate in the IPO.

Speculation, Investing, and Circles of Competence

I have been asked on many occasions why value investors tend to steer clear of technology companies even in cases where the investor has a significant understanding of technology.  In my case, several years of experience in software at the time surely provided the circle of competence required to evaluate eBay’s business model and to understand the technology involved.  The trouble was that my circle of competence did not extend to being capable of anticipating the emergence of entirely new industries that had no proven economic track record.  Even if I had spent the time to read eBay’s proxy filings, it is unlikely that I would have been able to confidently predict cash flows or to quantify downside risks.

Would it have been possible to develop a circle of competence that included the ability to forecast eBay’s cash flows?  It would be arrogant to assume that no one could perform such an analysis and perhaps there were some observers who even managed to predict eBay’s future growth with some precision.  However, it is safe to say that most buyers of eBay stock at the time were engaged in speculation rather than investing as defined by Benjamin Graham in Security Analysis:

An investment operation is one which, upon thorough analysis, promises safety of principal and a satisfactory return.  Operations not meeting these requirements are speculative.

Thorough analysis and an understanding of the business would not be enough without also being able to confidently demonstrate the safety of the investment and prospects for returns going forward.  Significant wealth has been lost by individuals who mistake a technical circle of competence with an investing circle of competence.  It is possible to understand a business extremely well and not be in a position to intelligently value a company.

Example:  The Software Industry

Economic Characteristics

Let’s take a closer look at one industry that is very exposed to changing technology.  The software industry has economic characteristics that often provide successful companies with outstanding margins and returns on equity.  Successful companies typically have strong moats that provide protection against entrants, at least in the short run.  This is because switching costs in software tend to be very high.  Particularly in the case of commercial software targeting business users, companies do not wish to change vendors frequently due to the cost of implementing a new system and training staff members.  Buyers of commercial software are also very risk averse.  Usually, decision makers seek to minimize “career risk” by selecting vendors with proven track records that are recommended by consultants.

Sources of Revenue

Most software companies have three primary sources of revenue:  Software licensing fees, maintenance fees, and service fees.  License fees are normally very high margin due to the low marginal costs of providing an additional copy of software to a new customer.  Once the research and development costs have been incurred to produce a software product, the marginal cost of production for additional copies is extremely low.  Maintenance fees are normally charged for technical support and upgrades and typically have solid margins, albeit not as high as licensing revenues.  Many software companies also provide services associated with their products.  Service revenues have much lower margins due to staffing costs, but such revenues can still be worthwhile.  The revenue mix of a software company is important when it comes to valuation.

Forecasting Cash Flows

The economic characteristics of an established software firm allow for building models to forecast future cash flows with some precision given that revenue mix and margins tend to change slowly over time, except during periods of disruptive change.  It is all too easy to build spreadsheets with revenue and earnings projections far into the future.  Such forecasts can incorporate various assumptions regarding overall revenue growth, revenue mix, and margins and can be aggressive or conservative when it comes to growth projections.  The cash flow can then be discounted at the analyst’s chosen discount rate to arrive at a present value.  If the stock can be purchased at less than the indicated present value, would this qualify as an investment operation as defined by Benjamin Graham?

Technological Change:  The Wildcard

The fatal flaw behind this type of valuation model is that disruptive changes can occur in technology and software can be heavily impacted by such changes.  The nature of the change is not always incremental.  Change can sometimes appear suddenly and can invalidate business models that worked well for many years.  While the barriers to new entrants in software can be high within a “steady state” environment where technology is relatively static, the same is not true when technology changes rapidly.  In such a scenario, new entrants can leapfrog established players in a very short timeframe.

We have seen several examples of such change in recent decades, but perhaps no change was as profound as widespread adoption of the internet during the mid to late 1990s.  Incumbent firms that failed to adapt did not lose their revenue sources overnight, but they experienced steady erosion in short order.  In the case of business buyers, “career risk” drives decision making for most managers  and will lead to business for incumbent players, but there are always trend setters who will give new entrants with superior technology an opportunity.  Such new entrants can quickly displace incumbents and become the new “standard”.  This cycle happens again and again in the process Joseph Schumpeter called “creative destruction”.

Many new entrants that appeared in the 1990s are now the established incumbents.  Some incumbent firms in the 1990s successfully remained incumbents by adapting and embracing the internet and other advances.  Which incumbent firms today with substantial brands and economic moats will adapt successfully over the next decade?  Will “cloud computing” displace traditional applications or is it a passing fad?

These are all questions that can radically alter the economics of the industry in the coming years and the fortunes of existing players, even those with powerful moats in today’s environment.

Adapting to Change

When examining incumbent firms with current economic moats, it is particularly important to determine whether they have a track record of investing in R&D and successfully navigating major shifts in technology in the past.  Has the company cut R&D spending in the economic downturn to limit damage to earnings or have investments been maintained?  Does the company have a culture where change is embraced rather than feared?  Do employees “live and breathe” technology?  Some of these questions are unquantifiable but are still critically important.

What About Start Ups?

If it is so important to evaluate track records, then how can anyone invest in a start up?  This is a very important question, but one that may be outside the scope of investing as defined by Graham.  There is certainly an important role for venture capital firms and others who invest in early stage companies that have technologies which could be game changers and end up disrupting incumbent firms.  The economy would suffer significantly without venture funding.  Perhaps a widely diversified group of such companies can be selected with the idea that most will fail but the few that succeed will make the overall endeavor worthwhile?  While this may be true, I would still argue that such operations are not within the category of “investing” as defined by Graham because they fail the test of providing adequate safety of principal and satisfactory returns through quantitative analysis.

It is certain that closely adhering to the principles of value investing will result in missing most if not all early stage opportunities.  Value investors who choose to invest in technology firms are best served by focusing on established incumbent firms with attractive valuations and a history of adapting to change rather than speculating on unproven ventures.  If the investor does choose to participate in unproven ventures, it should be done with full awareness that the operation does not meet Graham’s standard as an “investment”.  This does not mean that the endeavor is unwise or doomed to failure, only that the tools of value investing are not suitable for providing guidance on the decision.

Disclosures:  None.

The Value Investor’s Technology Dilemma
Tagged on: