How Telecom Is Becoming A Cyclical Industry, And What To Do About It

DRAFT.

Eli M. Noam[1]

Columbia University

May 5, 2002

Second DraftJune 28, 2002

This article analyzes the long-term lessons of the recent upturn and downturn in the telecommunications industry. It concludes that suchcyclicality will be an inherent part of the telecom sector in the future. To deal with thecyclicalsuch instabilities,, the most effective responses by companies and investors is to seek consolidation and cooperation. Hence, an oligopoly is likely to be the equilibrium market structure. This means that government, if seeking stabilization, will need to reassess its basic policy approach that has long been focused on the enabling of based on competition. And this, in turn, means that the structure of future network industry will look a lot more like the old telecom industry and less like the new internet..

1. The New Cyclicality

How fast things have changed. Only yesterday, the sky seemed to be the limit for the telecommunications industry. But Ttoday the electronics-based new media economy has become an old-style bust. Legacy is in. Balance sheets are in. . We need not listen anymore to the purveyors of hype about how bits play by different business rules than atoms, how the silicon economy is different from the carbon one, and how a P/E ratio need not have any E that stands for earnings, as long as it stands instead for electronics. Moore’s sunny law of exponential one-way progress has met the dismal science, and the two did not get along.

[add downturn numbers on investment, traffic, stock valuations, competitors, etc for 1990, 2000, and 2001/2]

[Insert W]

Telecom stock prices, using February 1996, the birthdate of the Telecommunications Act of 1996, as the benchmark of 100, rose from about 70 in January 1995 to 225 in March 2000, and then fell back to about 80 in March, 2002 (Standard and Poor’s in Alleman, 2002). Among new entrants, of 28 publicly traded competitive local exchange companies, 13 went bankrupt in 2001. [add earnings, investments and revenues]

In the first quarter of 2002, the total number of phone lines for Verizon dropped by 2.7%, for SBC by 3.6%, and for Bell South by 1.8%. In contrast, in the recession of the early 1990s the total number of phone lines had increased slightly (Stern, 2002). There are predictions of real telecom industry losses of $500 billion, much greater than for the savings and loan debacle of the late 1980s. As network firms sought survival rather than expansion, the telecom equipment manufacturing sector all but collapsed.

The “new” telecommunications industry is characterized by new technology, new applications, and maybe new market structures. It is also beset by a new set of instabilities and uncertainties that will become the theme of telecommunications debates in the next years, in the same way that the topic of competition had dominated the discussions of the 1990s. If the change is not merely a short-term correction but more fundamental in nature, managerial and policy assumptions need to be reviewed fundamentally, too.

Quite likely, the present downturn is only temporary and the industry will recover, though no at the hyper level of the bubble years. That, however, is not the real problem for the industry. It is not a one-time recovery from a one-time boom and bust. The main problem is that the telecom industry is entering a chronic pattern of volatility, with boom-bust patterns becoming a common occurrence rather than an aberration. Thus the telecommunications network environment is leaving linearity and entering volatility. A pattern of ups and downs may be emerging, a cycle.

Yet manye. pParticipants and observers of the industry miss or deny the emergence of cyclicality, sometimes interpretingviewing the term too literally and needlessly implying it to mean a regularity of ups and downs. patterns Most observers believe that the present downturn is merely a one-time accidentevent like a train crash, that things will return to their past stability because we have learned from past mistakes, and that we will not repeat them). This view is one of denial. True, we do not have much experience with volatility in telecom to make long-term predictions. And true, we are learning from the recent past. But if we if we analyze the drivers of the recent volatility, we mustthe sources of expansion and the forces of decline, conclude that they will be with us into the foreseeable future, and cyclicality with them, just as they are in some other industries unless, that is, we take steps that are at variance with telecom strategy and policy of two decades.This is neither likely nor necessary for the arguments that follow. A regular cycle would be predictable, and one could therefore deal with it much better.

Business cycles are not new, of course. The Bible tells us of the seven fat years in Egypt, followed by the seven lean years, and of Joseph who, engaged in what we would today call what would be described today as a counter-cyclical economic policy. Economic historians have identified more than 30 cycles for the years since 1854. There have been nine “official” recessions in the US since 1950.There have been nine “official” recessions in the US since 1950. Expansions have become gradually longer, contractions shorter, and milder (Zarnowitz, 1992). But they have not gone away. One reason is that we know better how to deal with downturns through macro-economic policies, and how to mitigate their negative effects through social policies. A dampening of the business cycle has been a priority for all governments.

Within the aggregate economy there are cycles for various asset classes and industries. The swings of the economy as a whole is a composite of the moves of its various sectors and firms. Just as in the case of investment portfolios, the aggregate volatility of the economy is lower than that of the average of the individual volatilities. These volatilities are interrelated in various reinforcing and offsetting ways. The swings of the economy as a whole is a composite of the moves of its various sectors and firms. Just as in the case of investment portfolios, the aggregate volatility of the overall economy should be lower than the average of the industry volatilities.

Some industries are more cyclical than others. One study of about 250 industries found that durable goods industries are three times as cyclical as non-durable goods industries. (Petersen and Strongin, 1996). But interestingly, unionization was not found to be statistically associated with cyclicality, neither as a cause nor as an effect, despite the greater downward inflexibility in cost it is reputed to create.

While business cycles are not new to many industries, in telecom they are an entirelya new phenomenon. In the past, the network industry progressed in only one direction: up. Telecom used to be less volatile than the economy as a whole. It grew steadily, with long planning horizons hardly ruffled by the business cycle. (The only time the industry declined in volume was in the Great Depression, when subscribers dropped temporarily by 16%). One company, AT&T, accounted for 80% of network activity and equipment manufacturing, and provided stability, planning, and an industry-wide umbrella. Its stocks with their steady dividends were treated by investors like bonds. The equipment industry, being also globally diversified, was almost as stable as the carriers. But today, in sharp contrast, it the telecom sector may well have become more volatile than the economy, more like the construction business, less than water utilities.

The telecom industry was unprepared for such a transition. Telephone lines had never declined in number before. The financial community was similarly lost, and its investment advice and valuations, assuming only growth were wide off the mark. Similarly, no academic literature had prepared us to the topic as applied to the telecom sectorvolatility. Being caught by surprise, aGiven this novelty, a massive hand-wringing and finger-pointing ensued.

This reaction is not altogether surprising. It is characteristic of industries (orand countries) that had previously experienced long periods of growth to be especially unprepared for a downturn and be frightened by it, because it is such a novel, almost existential experience. The South Korean national economy is one example. Telecommunications is now another.

2II. Why Cyclicality?

This question is important, because if we do not know why something happened, we cannot predict, prevent, or encourage recurrence. [2] when and how it will go away or re-appear.

There are many competing explanations for economic cyclicality, from sunspots to the alignment of the planets, and to the political election cycle. Over more than a century and a half, many distinguished economists have contributed their views[3]. I will discuss six main theories approaches and relate them to the telecom industry.

21.1 The Monetarist View. According to that viewtheory, associated especially with Friedrich von Hayek (1933, 1950) and Milton Friedman (1982), cycles are caused by flawed monetary policy that causes instability. For example, if a central bank changes interest rates incorrectly, consumers and businesses get wrong signals and their expectations lead to reactions that set off instabilities. (Fischer 1997).

Whether this theory is correct or not for the aggregate economy, it is probably not applicable to the telecom industry. Whatever telecom’s problems are, interest rates, even after being low and contributing to the bubble, did not rise to levels that would put them high on the list.

2.2 The Keynesian Perspective. Aggregate demand is affected by the mood swings of market participants which often become self-fulfilling. (Keynes 1936, Hicks 1950, Tobin 1975). The key trigger is psychological and, on the demand side. Keynes called it the “animal spirits” of entrepreneurs. Today it has been termed by More recently Allan Greenspan described it as an “irrational exuberance” in which the stampeding herd of stock market bulls eventually turned into a throng of lemmings (Shiller, 1978, 2001).

The demand orientation of the Keynesian approach leads Wall Street analysts to look closely at data for consumer spending as leading indicators. But for telecom and the internet, one cannot really blame a drop in consumer demand on the downturn. True, the growth rate in the internet subscribership are not as torrid as before – it was “only” 30-40% in 2001 – and internet minutes per user and day have dropped a bit, as one would expect as more marginal subscribers are signed up. Internet connections increased in America by about 20 million in 2001. Broadband internet subscribership is up and with it the aggregate bit flow for the internet. Similarly, the usage of long-distance minutes, data communications, and wireless minutes keeps rising. With double-digit growth rates in actual consumption, it is hard to blame insufficient demand on the downturn.

2.3. Real Business Cycles theory (RBC). This theory is a supply side story, going back to Prescott (1983) and others. For RBC advocates, cycles are caused by random shocks and their impact on total factor productivity. The internet was a positive shock. September 11 was a negative shock. Random positive shocks lead to higher productivity, higher output, higher real wages, consumption, etc. For RBC advocates, causality does not run from consumption to output but the other way around. (Espinosa-Vega and Guo, 2001). It therefore rejects explanations based on consumer psychology such as “exuberant irrationality.” RBC proponents believe that there is really nothing that governments can do about a cycle since it is based on random shocks.

How does this perspective fit telecommunications? Empirical studies show that single shocks do rarely trigger downturns. But a shock can topple an already weak structure. In telecommunications, several shocks occurred in the same period and added their impact cummulatively. Local competition failed; long distance competition, on the other hand, worked only too well, lowering prices and profits; the dot-com sector crashed back to Planet Earth; Wall Street became irrationally depressed when stocks declined from their unrealistic heights; governments extracted future expected profits by auctioning access to a vital resource, spectrum; spectrum auctions squeezed the industry; regulators, often beholden to incumbents’ well-being, held back competitors; etc. But note that most of these events are not the kind of exogenous, technology oriented, shocks that affect productivity, as hypothesized by RBC advocates. They are endogenous financial and institutional variables of the sector itself. They are thus not truly random but systemic.

There is, however, one important factor that can be interpreted as a technologically-based shock, if we take a generous definition of the term. It is the re-emergence of economies of scale of through network technologies such as fiber-optic distribution transmission cables and of wireless distribution systems. On the supply side, the fixed costs of networks are rising and the marginal cost are dropping – strengthening the classic attributes of “natural” monopoly. Scale effects are compounded by “indivisibilities” or “lumpiness” in investment, which leads to short-term excess capacity (Darby, 2002). Hence, the advantage of being large are greater than before. Size matters. As a result, for example, the market share of mobile wireless telecom providers (i.e., relative size) has been an excellent predictor of profitability (Waverman, 2002). And the cost of wireless firms is inversely related to the absolute and relative size of wireless companies, as exhibited in the graph below (Katz et al, 2002).

Similar effects have long been identified for the telecom long distance market (see, for example, Denny, Fuss, and Waverman (1981), Nadiri and Schankerman (1981), Alleman (1983)), as well as for cable TV (Noam 1983,1985). They generally show cost elasticities with respect to size of 5-15 %. In addition, the technological expansion in capacity has not only increased but accelerated (Noll, 2002).

[Rob: where is the 2^nd graph from Booz Allen repost? I faxed it.]

For a long time, these size advantages in telecom were submerged under the accumulated inefficiencies of the incumbents. But having had to shape up under the pressure of real or threatened competition, they reduced their inefficiencies sufficiently for their scale to overcome the threat efficiency of small entrants. This is not the classic RBC story, but it is inspired by it.

2.4. Lag and accelerator models. These models go back to Samuelson (1939). Small changes in desired capacity levels lead to large differences in capacity expansion, which drives investment. Where there is a delivery lag, unanticipated shifts in desired capacity can generate cycles of investment spending. The key here is the adjustment lag. These lags induce oscillation in the same way that a slowly reacting bathroom shower induces cycles of hot and cold water. The famous “cobweb” cycle is a model of such overshooting. Industry examples are cattle, airline services, and office space , where it takes a long time to increase capacity – and now telecommunications. Here, investments take a long time to get on line, and disinvestments may take even longer (it took more than a decade for the excess supply of 1980s Texas office space to dissipate.) The lumpiness of investments in telecommunications, coupled with an even slower regulatory and court system, makes the feedback loop very slow. On top of this is the “chicken and egg” problem of applications development depending on networked buildouts and vice versa. Since it is difficult to synchronize the two, developments often progress in spurts. The build-out of networks for broadband-internet capability is a recent example.

2.5. The “Austrian” theory. This view is associated with Mises (1928xx), Hayek (1933, 1990), Haberler (1937), Böhm-Bawerk (1895xx), Wicksell (1936), and Schumpeter (1939). It is focused on overcapacity. Assume Such overcapacity has been created for some reason—whether due to exuberance, excessive bank lending, monetary policy, or other factors. After an adjustment lag there is eventually a downturn. The pattern is one of boom, overcapacity, price war, bust, shakeout. A young industry tends to start off with small firms, and once their product fetches a high price it attracts entry, which expands output and lowers price. This goes on for a while. Industry growth rate slows below that of individual firms, and a shakeout occurs. For example, there used to be 275 tire manufacturers in US in 1922. Today, less than a dozen survive, even though the tire production as a whole is vastly larger. This view is common-sensical, with numerous examples such as snowmobiles, pocket calculators, bowling alleys, PCs, or movie theaters.

The Austrian view seems to describe well the telecom industry, in which the various network companies over-optimistically projected long distance market shares that added up to over twice the actual market. Everybody built capacity to overwhelm competitors and gain size. Capital expenditures grew by an annual rate of 29% from 1996-2001. The incremental cost of bandwidth fell by about 54% annually. Overcapacity was assisted by the lumpiness of telecom investments such as oceanic cables and its irreversabilities. It was further assisted by the tendency of Wall Street analysts to value a firm’s progress by physical measures of its infrastructure, such as cell-sites and fiber-miles. As the result of these factors, some carriers had over 90% of their fiber “dark” and prices dropped dramatically.

The Austrian theory has its detractors. Paul Krugman, the MIT Princeton economist and New York Times columnist, has called it a “hangover theory” and “about as worthy of serious study as the phlogiston theory of fire” (Krugman 1998), largely because it places the blame for a downturn in the boom itself, and does not explain why each company systematically over-estimates its market. Krugman’s critique is focused on the macro-economy rather than specific industries, where he accepts the notion of over-expansion. For telecommunications, then, one would need to understand why so many companies were so wrong about their prospects, and why none engaged in a meaningful counter-cyclical strategy.

Paul Samuelson, one of the world’s most distinguished economists, indirectly addressed this criticism, though he, too, is a Keynesian. Samuelson by observinged that economists have no theory to predict when a bubble in asset prices will burst. It is, therefore, not “irrationally exuberant” but rather perfectly economically logical rational for an individual to participate in a bubble. An individual’s micro behavior may be efficient even if the system is macro-inefficient.

2.6. Externalities. The RBC theory discussed earlier assumes constant returns to scale. That is, if one increases the capital and labor inputs of the firms proportionately, their outputs would grow by the same proportion. But for network industries this ignores the network effects, also known as positive the network externalities (Farmer and Guo, 1994) or the Metcalfe effect. An increase in usage leads to greater utility of the product and to increased demand. This increases productivity and real wages and enables further consumption. Growth of other network participants is factored in as part of the value of the product, and leads to still further growth. At some point, however, the expectations of further growth decline, for example as saturation occurs. This leads consumers to reassess the value of the service, and to adjust their consumption to the new value. This reduces demand or at least its growth, which creates negative network effects. And thus, the dynamics that had led the system to go up now take it down faster, too. Network externalities strengthen the oscillations of market demand rather than dampen it[4]. If the firms could coordinate their action they could jointly reduce this externality accelerator effects, but in a competitive environment they cannot easily or legally do so. This story fits well to the telecom and internet markets of the ‘90s and beyond.

2.7

Adding up the Theories of Cyclicality

Like the proverbial six blind observers of an elephant, each of these theories gets something right in its views of the drivers of cyclicality. Demand growth has slowed. Investment and regulatory lags prevented adjustment. Network externalities and lumpiness in investments amplified the swings. Economies of scale and network externalities created strong incentives for growth strategies, at the expense of profitability. Financial markets encouraged this strategy. Managers benefited from it in the short run. Yet while it expansion made sense for each firm individually, it created an a major oversupply in the aggregate.. Firms also overestimate the size of the market and of their share in it.

The following graph illustrates the increase in capacity for the North Atlantic market over a period of the past 3-4 years. The line rising to the right is capacity, and its increase is enormous. The declining line is the price. The line near hugging the bottom is the capacity required for trans-Atlantic voice traffic. . (To this one should must add capacity to deal with peak load periods, plus, of course, with data traffic, for which time series data of actual traffic is difficult to obtain. Data traffic is likely to be larger than voice, but not by the hugelarge multiples that would move actual usage much closer to capacity. (It is difficult to obtain the numbers for actual data traffic).

On the other hand, the multiplexing of voice reduces it from the assumed 64kbps for this calculation (Alleman, 2002). Thus Tthis graph is merely suggestive of the huge capacity overhang relative to voice traffic. This picturee increases in capacity and declines in prices would be even more dramatic if graphed back into the 1980s, or shown for trans-Pacific transmission, where prices of high-capacity circuits have dropped by 87% in two years.

In telecommunications, technological and economic obsolescence will gradually take capacity out of circulation. Satellites, for example, eventually leave their orbit or burn up. do not stay up forever, for example. But disinvestments takes time. For Texas office space, it took over a decade in the 1980s, to dissipate the excess supply. For railroads, it took many decades. Thus, the capacity overhang in telecom will remain for a long time if it is reduced merely by obsolescence. The key for a recovery is a substantial growth in demand, probably from mass-media uses of the internet such as video. When demand has caught up with supply, prices will rise, supply will expand, new entrants will emerge, and every firm will aim at increasing its market share. A new cycle emerges.

3III. The Implications of Cyclicality

Given this emerging cyclicality in the telecom market, what should be the response? I will discuss three types of participants: telecom managerscompanies, investors, and governments. The conclusion will be reached is that for the private sector participants the strongest strategy is to deal with cyclicality by seeking (or financing) an oligopolistic market structure. This, in turn, has implications for government policy.

Managers3.1 Telecom Companies

There is, of course, no shortage of potential actions for telecom managers companies to take. They includemust start with an intense :

Engage in self-analysis. Firms, managers and their owners and investors In the first instance, firms need to disentangle their owntheir firm’s performance from that of their industry, their customers and suppliers, and the regional, national, and global economies. This is not easy. But it is necessary in order to judge management’s performanceover the cycle,from the past. Following such analysis, several strategic options exist.

Cut cost and contract. A downturn puts more pressures on the firm and makes cost-cutting more acceptable. Similarly, a downturn provides an opportunity to change the internal structure and shed marginal operations. This strategy It may also include thes deferraling of innovation due to its riskiness. This strategy works best if competitors, too, follow it and slow the rate of their own investment in innovation.

Expand in the downturn. The opposite strategy from contraction may also make sense. The prices to acquire other firms through mergers or to expand by internal investment drop in a downturn, and it is a good time to get ready for the upturn, especially where lag times for investments are long.

However, in many countries the dominant telecom incumbent cannot easily be contrarian, because it may account for 80% or more of the market. This strategy might work better for a small firm—assuming it can raise the capital in a downturn, which is a fundamental dilemma.

One conclusion is that a strategy of expansion of capacity by internal growth makes less sense than the acquisition of a competitor’s capacity. Such a merger does not add to overall industry capacity (which would lower prices and profitability) but instead eliminates a competitor, which may result in higher returns.

Design the firm for downward (and upward) flexibility. Firms need to operate with built-in adjustments for the cycle. They need to implement scalable technology and flexible labor costs, for example, through profit sharing, commissions, and outsourcing. However, the labor component is becoming smaller as capital-labor ratios increase in the economy, and is therefore becoming less of a factor than it was in more labor intensive days. Similarly, firms need to engage in financial hedging. And in their capital structure, firms need to substitute corporate debt with convertible debt or preferred equity because this enables them to reduce a debt load in a downturn. In contrast, the capital structure of large telecom firms has become weighed by debt: it was 325% of market capitalization of France Telecom in 2001/2; 163% of Deutsche Telekom; 60% of BT; 66% of Telefonica.

Diversify in product markets and geography. Diversification reduces risk in some ways, but also may also get a firm to move outside its core area of competence, which raises risk again. Expansion into other countries creates exposure to political vagaries. Vertical expansion into related elements of the value-chain may create synergies (economies of scope) but may tie one part of the firm to others within the same corporate family, even if they are costlier and less desirable. It can also lead to a competition with one’s own customers. And, any expansion into multiple product lines inevitably creates and requires changes in the firm’s corporate culture, which may entail significant and mostly hidden costs.

Properly assess the risk of the business in a volatile environment, and factor it into all decisions.

Avoid a heavy debt load. In the downturn, cash is king. It reduces payment obligations and enables acquisitions. In the recent past, even established telecom firms have loaded up on debt in dramatic ways, and the result is that they are hurting in the downturn. Some of the reasons for such debt load was that the resource requirements of global and product expansion are huge, for example for the move to next generation wireless. Add to that the expense of firms engaging in empire building and the shortening of product cycles of technology, and the debt begins to balloon. Deutsche Telekom paid over $40 billion for the second-tier mobile carrier Voicestream; France Télécom paid $30 billion for Orange; and Vodafone paid $180 billion for Mannesmann.

After the 1996 Telecom Act a huge investment boom took place in the industry, in the order of $1.3 trillion. Between 1999-2001 alone, US telecom firms borrowed over $320 billion from banks. Credit was not difficult to obtain, and banks often accepted a subordinated creditor status without much collateral. But the revenues per investment dollar dropped. In 1996, according to Lehman Brothers, it was $5.08. But in 2001 it had fallen to $2.84 (Darby, 2002). Merril Lynch estimates that return on equity for the telecom industry declined from 13.8 % in 1996 to 5.9 % in 2001. In consequence, investments are expected to decline by an annual average rate of –14%. Verizon alone invested 33.7 billion over the period 1999-2002 with limited impact (Stern, 2002). Qwest’s debt load in 2002 was $25 billion. Around the world, the same story can be told. Deutsche Telekom’s debt is $64 billion, France Télécom’s $68 billion, and Telefonica $20 billion. As a percentage of revenues, France Télécom’s debt is 141%, Deutsche Telekom’s 140%, Telefonica’s 92%, BT’s 75%, and Telecom Italia’s 67%. Even these mountains of debt are being understated by various means. The cumulative debt of the seven largest European telecom firms exceeded $210 billion in 2002, greater the GDP of Belgium. It required some firms, on a daily basis, to commit over $10 million for debt service.

In theory, fAvoid a heavy debt load. In the downturn, cash is king. It reduces payment obligations and enables acquisitions. With the resource requirements of global and product expansion (e.g., the move to 3G wireless is hugely expensive) and a bit of empire building, even established telecom firms have loaded up on debt in dramatic ways, and are hurting in the downturn. But this must be qualified. The conventional view holds thatfirms without leverage will do better in downturn. But economy-wide studies also show that some highly leveraged firms have been helped in downturn by banks which cancelled payments rather than foreclosed. (Field, 1985, in Mascarenhas and Aaker 1989). and bySimilarly, governments that aremay be helpful in the downturn, especially when it affects several firms in an essential industry. Given such a safety net, big telecom firms had fewer reasons to be prudent. But they had an incentive to be too big and essential, to be permitted to fail, and this, encouraged expansion.

Similarly, [ADD]for its capital structure, firms may substitute corporate debt with convertible debt or preferred equity because this enables them to reduce a debt load in a downturnDeclare Bankruptcy. This step (e.g., a Chapter 11 reorganization) may wipe out debt, but will also make investors and lenders wary of future participation with the firm or its industry.

Engage in price cutting. This strategy has drawbacks when price cuts are matched by competitors. Hence, there are incentives to oligopolistic cooperation entailing the mutual reduction of capacity rather than engaging in price wars.

Diversify in product markets and geography. This reduces risk in some ways, but also may get a firm to move outside its core area of competence, which increases risk again. Expansion into other countries poses a problem of timing and exposure to political vagaries. Expansion into related elements of the value-chain create synergies but may tie one part of the firm to others within the same corporate family, even if they are costlier and less desirable. It can also lead to a competition with one’s own customers. And, any expansion into multiple product lines inevitably creates and requires changes in the firm’s corporate culture.

To conclude: many of the strategies in a downturn have common elements to make them successful: expansion and consolidation of surviving firms within the industry, and collaboration, to the extent possible, with one’s competitors.

3.2 Investors

In the past, the underlying assumption for government policy in telecommunications policy had been that if one gets a competitive structure in place, investment will follow efficiently and plentifully (Darby, 2002). The present downturn challenges negates this assumption in the short run (which is tolerable) and possibly in the long-run (which is much less tolerable).

For better orf for worse, there have been increasingly close linkages between financial markets and the real economy, with the links forged through such instruments as securitization, derivatives, and leveraged investments. Financial markets are also becoming increasingly volatile, as risk taking becomes easier but risk assessment harder. These trends, taken together, mean that financial turbulence makes industries more vulnerable to financial shocks.

Wall Street has thus been part of the problem, though it has also paid dearly for it, in money as well as credibility. Share prices seem to move in ways that seem unrelated to the underlying discounted value of cash flows. In fairness, it is hard to value companies in volatile industries, both conceptually and institutionally. “Consensus” earnings forecasts have been useless for investors in cyclical industries, especially for startup firms in high-growth, high-risk industries. Forecasts tend to be rosy and asymetrical. . According to a McKinsey study, for volatile industries the forecasts are generally upbeat, and “the forecasts don’t acknowledge even the existence of a cycle” (de Heer and Koller, 2000). Why this positive bias? Maybe it is the pressure of the investment banking part of the firms to avoid unfavorable evaluations of companies which would lose them as clients. Maybe it is the opaque accounting practices, and proliferating derivativeng securities and stock options that make assessments difficult. The authors sum up: “In light of these worries, it is reasonable to conclude that analysts as a group are unable or unwilling to predict the business cycle for these companies.” [p.65,66].

According to Joseph Stiglitz (2002), deregulations are often always associated with periods of frenzied activity that often tend to go wrong. And indeed, after the 1996 Telecommunications Act, financial markets gave the wrong signals to firms by raising stock values enormously. They evaluated unprofitable startup firms by proxies such as fiber miles and cellsites. This led telecom firms’ to over-invest in such physical elements,tment and to an over-investment in such firms. The bubble was further fed by the ability to borrow against the rising value of the stock; later, when of course as stock values dropped, they this necessitated further sell-offs and led to further price declines, etc. Traditional theory has it that investors look at total risk-market risk plus financial risk. If they are additive, then high market risk in this sector should have been offset by investor unwillingness to support highly leveraged capital structures. But, it seems just the opposite happened. Eventually,

What triggers the downturn are rapidrRapid changes in market expectations that lead to rapid reassessment of asset prices. This causeds massive reduction in asset-holder’s wealth, which lowereds expectations further. Clearly, returns were retreating broadly, and investments with them. As that happened, zooming share prices ceased to be a motivation for investments, and the new entrants rapidly fell out of favor, since they could not offer any earnings or service. Nor could they source their debt. Confidence was further shaken by questionable accounting practices.

Studies show that the inability of lenders like banks to discover the relevant characteristics of borrowers projects, e.g., when information is asymmetrical, leads to poor projects driving out good projects. This is Akerlof’s “lemon” principle (1970). In the new media and telecom sectors, lenders’ and investors’ ability to evaluate projects has declined, and with it the quality of loans. When this became obvious even to the investment bankers, they raised the threshold, cutting off in the process projects they could evaluate before.

For a firm’s stability, in a volatile environment, stock is better than debt. Since it does not require interest payment in a downturn situation. The increasingly high debt load of telecom firms used to finance various ventures thus subject these firms to more stress. Now the real prospect of default sends shudders through the financial community and leads it to support incumbents’ and their profitability against new-style entrants.

Investors now favored incumbent telecom firms, which looked much better than before when they had been derided as “dinosaurs” by the financial community. In contrast, the financing of new entrants largely dried up. If anything, further stability was desired. The incumbents’ traditional shareholders seek a utility-style stock with predictably steady dividends. When the industry becomes volatile, such investors leave. Market power, on the other hand, lowers risk, and raises prices and cash flows. One can see how market power benefits market valuations. In America, rural LECs, facing little prospect of competition, maintained their value much better than other telecom firms. In other countries, Telmex did better than most large incumbents, mostly because of its hold over the Mexican market. Its stock rose 28% in the first four months of 2002.

One factor for the mispricing may have been that the value of “real options’ was not properly accounted for. The opening of a market to competition, for example, creates an option to enter. These options are not exercised in a smooth pattern, and can sit idle for a long time and then take off in a burst (Grenadier, 1996). The valuation of these options is important because under existing practices, they are not accounted for, and hence are missing in the financial evaluations of firms and industry (Alleman and Noam, 2000).Traditional theory has it that investors look at total risk-market risk plus financial risk. If they are additive, then high market risk in this sector should have been offset by investor unwillingness to support highly leveraged capital structures. But, it seems just the opposite happened.

The downturn has also cast doubts on the Efficient market hypothesis Efficient Market Theory. According to that hypothesis, which has been a cornerstone of finance theory, (see, e.g., Fama 1970) financial prices reflect all available public information and hence are always correctly priced. Market trading should eliminate mispricing. Yet it is not easy to reconcile this view convincingly with the extreme swings in stock prices in the telecom market. Obviously, new information has become available. But much of it – on excess capacity, commodification, economies of scale – had been widely available before.[ADD] It was not information that had been missing, but its proper analysis.

3.3 Government

The wisest thing for government would be to ride out the cycle, but a hands-off policy might not be easy to maintain. Cyclicality is undesirable to government. When it comes to Schumperterian dynamics of “creative destruction”, governments dislike destruction even more than they fear creativity. Clearly, too much stability is also undesirable. See the example of Japan. What is the proper level of instability? Too little of it reduces innovation, but too much stability does the same. An optimum instability might exist, but it is difficult to agree on that level. Entire political philosophies hinge on the different views on the acceptable societal risk.

To the extent that volatility raises uncertainty it also raises the cost of producing telecom services, which is an essential and universal input. This has also some distributional implications such as fluctuations in employment. And through network effects, everybody is negatively affected. The losses from cyclicality can be substantial. Estimates for the economic losses from the oversupply in US of office space in the 1980s are $130 billion, in lost rents only, without counting the secondary effects of reduced tax receipts and negative multiplier impacts on the rest of the economy.

Imagine the impact of a telecom downturn on smaller countries where a telecom firm has a large presence and is affected by shocks originating from the outside. In Finland, Nokia accounts for 35% of exports and 14% of GNP. Similarly, the telecom sectors of less developed countries have become more volatile as they have opened up to the rest of the world and became engulfed in external instabilities over which they have no control.

All of this and more are reasons for government to fear cyclicality and to engage in counter-vailing policies, even though government policy may also be a contributing cause to cyclicality, for example through regulatory delay. If such involvement is likely, for better or for worse, what are the potential tools of government for dampening the cycles of the telecommunications industry? Some such tools are discussed in the following and they should be read, as a list of potential actions rather than as recommendations.

Flexibility in taxes and other payments..

Such a policy tends The US telecom industry is subject, to telecom taxes beyond the the regular business taxes. Those taxes could be automatically adjusted through the cycle if the tax rates was levied on earnings rather than on revenues. Similarly, spectrum license auctions could be structured in a way that would collect payments over the life of the license, with annual collections based on earnings, and would thus become thus an a