The impact of the Efficient Markets Hypothesis on damages
Damages: real or illusory?
This paper sets out the facts of a hypothetical case and discusses the damages issues raised by it. This paper should not be construed as expressing opinions on matters of law, which are outside the scope of the author's expertise. Nor does this paper represent the view of FTI Consulting Inc or any of its experts, who have held a range of views on the matters discussed below and may be expected to do so in future.
In August 2008, Düsseldorf & Dortmund Holding (D&D), a consolidated vehicle manufacturer, agreed to sell a stake in its European truck manufacturing business to a Chinese investor, Guanzhou Big & Heavy (GBH). D&D would contribute the assets of its European truck manufacturing operations into the venture, including human resources and intellectual property, in exchange for 50 per cent of the equity in a new joint venture company. GBH was expected to contribute e100 million in exchange for the remaining 50 per cent of the venture, to be known as Glorious Life Term (GLT). The transaction was set to close around seven months later on 27 February 2009, the last business day of the month.
Between the signing of the contract and the closing of the deal, much work was done. New entities were created and the assets transferred to them. GLT's systems were separated from those of its former parent. Employee contracts were redrafted, once it was known which employees would leave D&D to join the GLT joint venture. Standalone support functions, such as finance, human resources and facilities management needed to be created from scratch. Nevertheless, the project was well managed and as the closing date loomed, D&D felt that it had done its best to comply with its obligations.
Around two weeks before the expected closing date the chairman of GBH, Li Feng, rang his counterpart at D&D, Gottfried Heinrich, late one evening. The news was bad, said Mr Feng. Although his board had previously given approval to proceed with the joint venture, the provincial authorities, who owned a golden share in the GBH business, had withdrawn their support for the venture. Mr Feng stressed that, although he personally remained in favour of the deal, he had been instructed that GBH must refrain from any further pursuit of the joint venture. If he refused to comply, Mr Feng said, he might be personally subject to criminal proceedings. For that reason, he regretted that GBH would have to withdraw from the transaction. Mr Heinrich, clearly shaken, immediately called his legal team. The following week, D&D filed for arbitration against GBH for breach of contract.
The case for damages
The main claim for loss advanced by counsel for D&D was for the 'lost bargain' implicit in the joint venture agreement.1 In between the signing of the joint venture agreement and GBH's withdrawal in February 2009, markets around the world had declined precipitously in the wake of the collapse of Lehman Brothers. The deal price, however, was agreed at a time before the full onset of what appeared to many investors as an unprecedented global financial crisis.
The foundation of D&D's claim for loss of a bargain was that GBH had pulled out of the joint venture because it believed that its 50 per cent stake in the business was no longer worth e100 million. In the view of D&D's counsel, the difference between the contractually agreed price and the (lower) value of the assets at the date of GBH's withdrawal represented the true extent of D&D's loss. D&D claimed e41.7 million in damages, reflecting the fall in the Dow Jones Automotive Index (INAA.DE) between 1 August 2008 (25.54) and 27 February 2009 (14.90).
In its statement of claim, D&D relied on the Efficient Markets Hypothesis (EMH) and an assumption that its truck-making business was representative of the sector tracked by the German automotive index. D&D's expert, Jean Lannes of Investisseurs Rationnels Sàrl, said that:
The Efficient Markets Hypothesis, which asserts that market prices are always indicative of fair value, is a cornerstone of modern finance and a sound basis on which to assess any change in the value of the joint venture's assets."
In its reply, GBH simply stated that it had been prevented by unnamed members of the provincial economic council, which exercised the rights conferred by the government's golden share, from proceeding with the transaction. It protested that it would have wished to pursue the transaction and expressed regret that the deal was no longer on the table today. Counsel for GBH pointed out that, since the deal fell through, the German automotive index had since recovered to reach new highs in excess of its level in August 2008, when the joint venture agreement was signed. For that reason, GBH argued that any damages at all would unjustly enrich the Claimant, which still retained 100 per cent of its (recently quite profitable) truck manufacturing business.
GBH's expert, Stephen Austin of Animal Spirits LLP, prepared a discounted cash flow (DCF) valuation model of the proposed joint venture truck manufacturing business. The model valued the business using projections made available to GBH at the time of the transaction, discounted by a cost of capital independently assessed by Mr Austin.
In his report, Mr Austin stated that although the market risk premium probably rose during the period of the financial crisis, risk-free rates fell, largely cancelling out any increase in expected risk. In consequence, his assessment of GLT's cost of capital remained constant throughout. He also noted that by 1 February 2012, the index relied upon by Mr Lannes stood at 30.74, suggesting, based on Mr Lannes' logic, that the business was today worth e120.4 million. Mr Austin concluded that:
The value of 50 per cent of GLT's shares at the date of signature of the joint venture agreement, both at the date of the alleged breach and today was, and is, comfortably in excess of e100 million… Whatever irrational gyrations financial markets may have performed in late 2008 and early 2009, these had no impact on the long-term fundamentals driving the profitability of D&D's truck division over the 30 or more years of the planned joint venture. I therefore conclude that the value of GLT remained unaffected, aside from a temporary hiccup in sales around that time.
In its second memorial, D&D:
- conceded the first of Mr Austin's valuations, agreeing that at the date of signature of the joint venture, the price agreed by the parties in arm's length negotiations represented fair value;
- maintained that its own expert's valuation at the date of GBH's withdrawal was fair and that the EMH was an appropriate basis from which to infer the fall in value of its truck division between the date of signature of the agreement and the alleged date of breach. Using Mr Austin's model and what he considered a more appropriate (and higher) cost of capital, Mr Lannes was able to calculate a valuation in February 2009 similar to the one implied by the decline in the market index. D&D asserted that, even if the loss was not crystallised at the date of breach, that made it no less real. Indeed, if it was a necessary test that the loss must be crystallised, the effect would be to encourage parties like GBH to renegotiate or breach contracts with impunity on any unfavourable turn of events; and
- noted, of the third valuation, that while the sector had in general recovered, the golden age of European truck making was now firmly over and, whatever short-term gains it might enjoy, D&D expected to lose ground in future to Far Eastern competitors. For that reason, it saw less reason to be optimistic about the business today than at the date of signature of the joint venture agreement. D&D concluded that the value today might be lower than it was in August 2008.
GBH's position remained unchanged in its final submissions. There was, it said, no loss of bargain. Although in good times, markets might be a useful indicator of fair value, it would be unsafe to rely on the unstable emotions driving markets in the months after Lehman Brothers' collapse. Finally, GBH said that the fairest way to consider damages was by reference to the value of the truck division today, given that D&D had enjoyed and continued to enjoy profits from the firm, possibly in excess of those expected at the time of signature of the joint venture agreement.
Market value and damages
One of the damages issues confronting the Tribunal is the appropriate date at which the loss should be measured. Clearly, if one is to adopt a valuation that takes into account market prices, that valuation will be sensitive to the date of measurement. Although the appropriate date of valuation is a matter of fact for the Tribunal, the selection of one date rather than another might lead the Tribunal to prefer one expert's assessment of damages over another and is therefore potentially a matter of great sensitivity.
The volatility in financial markets in recent years has meant that valuation conclusions based on financial market evidence can vary sharply with the date of measurement. For example, losses calculated on the basis of financial market data on 1 September 2008 (pre-Lehman), might differ by millions of dollars from those calculated using the same dataset a month later (post-Lehman).
It is not clear whether a Tribunal considering a case like D&D v GBH in 2012 should adhere to the pre-crisis asset price (reflecting the agreement between the parties), the post-crisis asset price (reflecting the impact of the crisis on value) or a later price that indicates whether the asset subsequently appears to have recovered in value (but which may not sit easily with the date of breach). When markets are calm, it might be easier for experts to agree that they are good indicators of fair value and for DCF valuations closely to approximate to market valuations. Outside of such periods, tribunals must choose their dates with greater care and, perhaps, with an eye as to whether the outcome implied by selecting a particular date is consistent with what it considers a 'fair' outcome.
In my view, however, tribunals should not fall into the trap set by Mr Austin. It seems likely that Mr Austin's DCF model shows an unchanged value throughout because it does not accurately reflect the drivers of business value. The reason why market values fluctuate is because they reflect changes in the 'consensus DCF model' adopted by investors. As market expectations change, whether about cost of capital, future economic growth, raw materials costs or wage inflation, so too do market values. A DCF model that does not reflect even the direction of changes in market value is unlikely accurately to have captured and modelled the factors that affect business value.
The strain in global financial markets in recent years has led some to question the appropriateness of using financial market data as the basis for measuring value in some, perhaps most, market conditions. There is a range of studies of 'market anomalies' which show that deviations from what might be termed 'fundamental' or 'real' long-term value may persist in markets for periods of months or even years.
There is also the issue of 'asset bubbles', in which asset prices are thought to be inflated above their 'intrinsic' value. Although some economists deny that bubbles occur at all, others are persuaded that financial history contains numerous examples of overpriced asset classes that subsequently rewarded investors poorly. Commonly cited examples include Dutch tulips (1630s), British railway stocks (1840s), dotcom shares (1990s) and Irish residential property (2000s).
If market prices of individual assets, or even whole asset classes, can 'undershoot' or 'overshoot' their 'fair value' for (possibly extended) periods of time, practitioners could more readily disagree about whether market value closely resembled 'fair value' at any given point in time. If 'fair value' could deviate from market value it might be be more appropriate to draw conclusions about value from non-market sources (such as DCF modelling).
The debate about market value is not confined to valuers and finance practitioners. There is also, for instance, a continuing debate in accounting circles as to whether the value of companies' assets should be 'marked to market' using financial market values at a given date. The mark to market debate is of particular importance to banks and insurance companies, whose very solvency may be threatened by unexpected and rapid changes in the market value of their assets.
There are three main potential pitfalls in using market data to value assets:
- market values can be volatile. While short-term volatility clearly matters to an investor seeking to trade in the market, it may be of less relevance to an investor who intends to hold the asset for a long period of time;
- many assets on companies' balance sheets are customised, such that there is no close proxy for them on any financial exchange. In these cases, judgment must be used to estimate the value of the customised asset, perhaps anchored to a market-traded proxy for the asset; and
- the accuracy and relevance of market valuations at a given point in time is sometimes questioned by certain academics and popular commentators. The debate about the accuracy of market valuations extends not just to small, emerging market stock exchanges but also to many of the companies listed on the largest financial markets in the developed world.
Members of the expert community differ as to the appropriate weight to award to financial market information in the assessment of damages. While there are those who believe that financial markets are 'always right', and therefore represent a sound foundation on their own for the assessment of value, others might consider that financial market prices are only one of a range of potential indicators of value. Sometimes, these differences between experts are a function of their views as to the accuracy of the EMH.
An outline of the EMH
What do we mean by an 'efficient market'?
A financial market, like a medieval cloth fair, exists to serve the needs of those who seek to convert cash into assets (or goods) and vice-versa. On a fundamental level, a properly functioning (or 'efficient') market would be expected to have a number of characteristics, such as:
- timely and accurate information about past transaction prices and volumes;
- high liquidity, the ability to buy or sell an asset quickly at a known price. The arrival of a new buyer or seller in the market should not move the price;
- absent new information, prices should not change from one transaction to the next, a feature known as 'price continuity';
- together,liquidity and price continuity are supported by 'depth', the presence of many potential buyers and sellers who enter and leave the market in response to changes in supply or demand, given their independent assessments of value;
- low transaction costs (as a proportion of the value of the trade); and
- prices that respond quickly to new information affecting supply or demand (ie, investors have 'rational expectations' about the future given their knowledge today). Because news is not able to be predicted, neither are stock prices.
Clearly, the larger volume securities traded on financial markets like the New York or Hong Kong stock exchanges, the Chicago Mercantile Exchange or Liffe meet all of the criteria above. Most of the world's securities and financial markets, however, do not because:
- there are only a few securities listed on the exchange;
- some or all of the securities that are listed are thinly traded;
- information is difficult to obtain;
- there are insufficient traders in the market for a given security;
- the market is split in some way, such that certain securities (or investors) trade on one 'board' of the exchange, while others trade on a different 'board'; and/or
- different market participants face different transaction costs or taxes that limit their willingness to trade in comparison to other market participants.
In reality, all markets are to some degree imperfect. Even on the NYSE, retail investors face different transaction costs and taxes from those of mutual funds. On any exchange, some stocks will be less frequently traded, potentially leading to wider bid-ask spreads and price movements even for relatively small trades. The issue is whether these imperfections significantly distort the price signals given by the market. In the case of large, developed markets and large cap stocks, there is likely to be no significant distortion. For minor markets or for illiquid stocks, however, market prices may need to be treated with greater caution.
What is the EMH?
The Efficient Markets Hypothesis states that all available information about a security is reflected in its current price. Implicitly, the EMH assumes the existence of an 'efficient' market with the characteristics set out above.
There is no single EMH. Rather, there are three sub-hypotheses about efficient markets, which have been categorised as:2
Weak-Form EMH: current stock prices fully reflect all security market information, such as historical volumes, prices, rates of returns and large-size transactions. The implication of the Weak-Form EMH is that all systematic information is embodied in the price, so one cannot earn superior returns by following a trading rule based on past rates of return or market data;
Semistrong-Form EMH: security prices fully reflect all public information about the security including both market information (as in the Weak-Form EMH) and non-market information such as earnings, earnings ratios, dividends, stock splits, competitor announcements and economic or political news. Once information is public, therefore, prices should quickly adjust so that investors are not able to make superior profits. Investors in possession of non-public information about the security, however, may still earn superior returns (subject to legal constraints); and
Strong-Form EMH: security prices fully reflect all public and nonpublic information. The implication of the Strong-Form EMH is that no group of investors has access to superior price-setting information and so no group of investors can consistently earn superior returns. In the Strong-Form EMH, markets are not just 'efficient': they are perfect because all investors are assumed to have cost-free and simultaneous access to information.
Over the years, many tests of the EMH (in its various forms) have taken place. A crude summary of those investigations is that:
- there is a broad range of evidence to support the Weak-Form EMH and a series of studies have confirmed that trading rules cannot systematically beat the market, net of transaction costs;
- there is mixed evidence to support the Semistrong-Form EMH. Although prices appear to react fully to news events, corporate events or accounting changes, several studies have found a number of pricing anomalies (eg, repeatedly high returns in the month of January) that would not be expected under this version of the hypothesis; 3 and
- there is also mixed evidence to support the Strong-Form EMH. Some studies have indicated that corporate insiders and stock exchange 'specialists' (a group of large market makers on the NYSE) do experience superior returns (in defiance of the theory), while others indicate that security analysts and professional money managers do not (in support of it).
A description of the evidence is outside the scope of this paper. I note, however, that the issues associated with robust testing of the EMH are neither simple nor clear. For example, it can be difficult to know what investors' expectations really were at any given date in the past and, therefore, whether prices were appropriately set at that time. Overall, however, it appears that violations of the EMH do occur, but that there is a substantial body of evidence in support, particularly of the Weak-Form EMH.
The EMH has been highly influential in financial theory in the last 40 or 50 years and forms part of what I shall term 'standard' financial theory today. Recently, it has come under attack from supporters of a group of theories known as 'behavioural finance', which have attracted interest, in part, as a result of the failure of 'standard' financial theory to account for recent observed market behaviour. For instance, in the second half of 2008, financial market swings occurred several times in a week that according to 'standard' theory should occur only once in thousands of years. Outcomes such as these might appear to be at odds with a premise of market 'efficiency'.
What might be wrong with the EMH?
Behavioural finance refers to a collection of theories that seek to understand how psychological and economic principles interact to affect financial market outcomes.
One example of the psychological effects that behavioural finance studies is heuristic bias, which lead people to draw incorrect conclusions based on the 'rules of thumb' that they have developed from experience and from which they draw inferences. For example, even though they are routinely told that historical returns are not a guide to the future performance of financial products, people, on average, tend to extrapolate from past performance of securities when considering their likely future performance. People are also prone to 'anchoring', in which they ascribe too much weight to their own past experience and so fail to react appropriately to new information.
A further example is frame dependence, people's tendency to choose form over substance. For example, faced with a (paper) loss of 10 per cent on a stock they own, people have a tendency to hold on to it in the hope that the stock will return to break-even terms. A more rational strategy, however, might be to crystallise the loss and invest in another security they think will perform better in future.
These psychological biases appear to be quite strong in many people, to the point where they may influence financial market outcomes through 'herding' behaviour. For instance, it has been shown that 'value investing', buying stocks that have generally performed poorly and so appear 'cheap' today, provides superior risk-adjusted returns to buying 'growth' stocks, which have generally experienced good past performance. It may be inferred that people neglect value stocks and pay too much attention to 'star performers' whose strong recent growth and high valuation multiples they expect, by extrapolation, to continue in future.
Other studies have shown that individual investors generally display excessive optimism (they don't buy enough insurance, especially life insurance), are overconfident (they trade too often, despite being at an informational disadvantage to professional investors) and discount diversification (they tend to pick too few stocks, hold too much of their employer's stock, focus too narrowly on domestic stocks and split their money naively between the available choices).
Collectively, these imperfections in investor behaviour drive patterns of returns at the market level that are sub-optimal and which have been observed to defy the accurate pricing of securities implicit in the EMH.
Is behavioural finance helpful?
In general terms, followers of behavioural finance assert that in the 'real world' investors are not always 'rational' (ie, profit-maximising and able to assess risk and return well). The implication is that markets are not always technically efficient, as explained above. Clearly, therefore, behavioural finance is at odds with the EMH. The question is whether this matters for the purposes of valuation and, if it does, when and why?
The first point to note about behavioural finance is that the subject is still evolving. Much research is continuing into the links between human psychology and market behaviour. There is no 'general theory' of behavioural finance that can be used by investors in the real world to guide them in selecting assets for their portfolios or strategies for their trading. Critics of behavioural finance sometimes observe that even if 'standard' financial theory is sometimes wrong, the new theories do not offer anything to replace it with.
The ideas underpinning behavioural finance may help to explain past market events that have fallen outside the boundaries of probability according to 'standard' financial theory. Even here, however, behavioural finance does not purport to offer a detailed explanation of (say) the extreme movements of financial markets following the collapse of Lehman Brothers. Rather, it is observed that they were caused by irrational behaviours, such as fear or herding instincts.
Equally, while the EMH can be misinterpreted to assert that 'the market is always right', the principles of behavioural finance are equally open to misinterpretation. Recently, the claimed 'inefficiency' of markets has been used as a pretext for state control and regulatory intervention. Whatever its claimed benefits, a ban on short selling will do nothing to make markets more efficient.
An assumption that 'irrational actors' influence market outcomes suggests that intelligent investors ('the smart money') should place their money differently, even contrarily, to the irrational actors ('the dumb money'). On this view, the price of a security on any given day may reflect transitory supply and demand between the smart money and the dumb money rather than an all-weather guide to fair value based on the latest information. Conclusions such as these make behavioural finance unsettling for 'standard' financial theory, but do not necessarily imply that behavioural finance is a superior guide to placing an accurate valuation on an asset.
Behavioural finance offers a number of painful truths to 'standard' financial theory. Clearly, 'standard' theory sometimes does not sit well with certain observed market behaviour. The problem with behavioural finance is that while it says why 'standard' financial theory may be wrong, it offers no alternative methods for assessing the value of assets in financial markets. To offer a valuation opinion today, a practitioner is still obliged to rely on theories, such as the Capital Asset Pricing Model (CAPM, used to calculate the cost of capital), which assume that investors are rational and markets efficient.
It is likely that most economists and finance practitioners largely agree that share prices are determined by rational agents in a competitive market using available public information. Many continue to concur with the stronger suggestion that stock prices reflect the 'fair value' of companies. Yet there is strong evidence not only that share prices are more volatile than would be expected given the price-influencing 'news' that emerges over time and also that share prices may be above or below their 'fair value' for extended periods, displaying irrational exuberance or pessimism and unbalancing the trade-off between risk and return.
Today, behavioural finance is perhaps best viewed as complementary to 'standard' financial theory rather than an alternative to it. Certainly, few in the industry have concluded from behavioural finance that the weight of investor opinion is so irrational that price discovery in markets has no informational content. The debate about how strictly EMH should be applied and, therefore, the weight that should be given to market prices in the assessment of value will therefore continue.
In the end, the best advice is to use more than one valuation method and then to seek to reconcile any differences in the resulting numbers. While neither 'standard' financial theory nor behavioural finance creates certainty about the fair value of an asset, the process of comparison and challenge obliges practitioners to reason, investigate and justify their conclusions about value.
In the case of D&D v GBH, such a process might have led the Tribunal to the conclusion that, although the value of D&D's truck business fell, as suggested by Mr Lannes' market-based analysis, a valuation based strictly on the market index might contain an element of 'irrational pessimism' (suggesting that Mr Austin's views about long-term value had some merit).
- In addition, D&D filed subsidiary claims for wasted expenses and the marginal cost of the finance raised on usurious terms to replace the e100 million withheld by GBH. Beyond noting their existence, the article does not discuss these further claims.
- See, for instance, Brealey, Myers & Allen, Principles of Corporate Finance, McGraw-Hill, 2008, International Edition, 9th Edition, Chapter 14, p359.
- Market anomalies, once identified, tend to disappear as traders arbitrage them away. The dominance of program trading and large institutional investors in today's markets may mean that there are fewer market anomalies today than in the past.