My puzzlement isn’t the product of some simple faith that markets correct all ills. It results from the difficulty of finding a theory that explains how market manipulation would work. To get a sense of why this is difficult, consider a simple information‐​related malady: person X lying to person Y. Of course, lying happens, but superficially it is puzzling because it works only if Y believes X is telling the truth. The economics of strategic behavior typically requires observed behavior to match expectations—“Nash equilibrium,” for jargon aficionados. Because it relies on mistakes, lying is “not Nash.”

One cannot solve this problem by invoking “asymmetric information.” As George Akerlof observed in his Nobel‐​winning article nearly 50 years ago, the only Nash equilibrium is for everyone to assume the worst. It is not that some get ripped off by assuming better and being disappointed; everyone should assume the worst and thus avoid getting ripped off. But being ripped off is just what a theory of market manipulation needs to explain.

As Penn State economics professor (and—disclaimer alert—friend and occasional collaborator) Andrew Kleit observes in his cogent, accessible, and informative book Modern Energy Market Manipulation, the first step is to define what one means by “market manipulation.” He begins by taking care to distinguish market manipulation from speculation or hedging. Like most economists, he recognizes that speculation and hedging are good for an economy. The former brings more information into markets and improves the value of prices as signals of expected future worth, and the latter minimizes the cost of uncertainty by exploiting negative correlations to reduce or eliminate risk. But Kleit goes further by identifying the focus of market manipulation not just as harmful practices but as those designed to change price rather than arbitrage price differentials or exploit price correlations.

Manipulations / Some practices that might fall under “market manipulation” in this sense are not mysterious. Cornering supplies of a resource, input, or complement drives up prices and is at the core of antitrust complaints based on “raising rivals’ costs.” Kleit wants to focus on prices of financial assets, however, and offers four candidates:

• Taking a long position in an asset and then releasing “allegedly factual information that increases ‘artificially’ the price of the asset.” An example of this that comes to my mind is the artificial orange crop forecast used to bankrupt the bad guys in the Eddie Murphy movie Trading Places. This implies that the manipulator has a monopoly over the information that influences the asset price. Otherwise, it is not conceptually distinguishable from Kleit’s next category.
• Reporting false trades that influence the price of an index that determines the settlement price of an asset. This comes closes to the “lying” story above. Kleit reports that indexes settled this way are now “rare” following market “learning,” although they may have been present in early electricity and natural gas commodity markets.
• Withholding supplies purchased in advance to drive up the spot price. This seems the easiest type of manipulation. But Kleit points out that this strategy requires that one first corner futures contracts without driving up their price, and then liquidate those contracts at some point without causing the price to plummet.
• Driving up the contract price by buying up assets during the settlement period. This seems to me to combine the previous two strategies. Kleit notes that this requires surprise because “no one will go short if they think the [commodity] price will be manipulated upwards.” But, as noted above, it is the requirement of surprise that makes market manipulation mysterious. Is it reasonable to think that those who invest substantial amounts in commodity futures markets are susceptible to surprise?

Kleit differentiates his definitions from those of Shaun Ledgerwood and others (second disclaimer: I’m an academic adviser to the Brattle Group, where Ledgerwood is a principal) in not requiring that manipulation activity be unprofitable but for gains from manipulation. This is surely right, but as with predatory pricing, a requirement that such activity be below the alleged perpetrator’s cost may be a useful rule to avoid chilling efficient conduct when enforcement is unlikely to be error‐​free. A below‐​cost pricing rule could also help distinguish manipulation from generally beneficial speculation or hedging. But Kleit, following decisions in leading manipulation cases, seems to prefer to use the intent of the alleged manipulator to sort bad conduct from good. However, he finds quantitative tests to challenge allegations in manipulation that were more likely hedging or speculation.

Kleit reviews the apparently limited theoretical literature on manipulation, but those models don’t offer much help. Two similar models, based on whether the manipulator forces down the price through selling contracts or the commodity itself in the spot market, appear to assume a manipulative effect of these sales on prices without explaining how that effect arises. He then models Ledgerwood’s characterization of index manipulation, finding that it works only if, in effect, the manipulator owns more of the asset than is used to set an index price—Kleit calls this a “direct effect”—or if the manipulated price changes the price of other transactions that set the index—called a “cascade effect.” He argues that the existence of cascade effects is uncertain at best and that only traders capable of surprise would use an index where a manipulator could profit from a direct effect. Because surprise is just the way normal people say that realized behavior differs from expected behavior, the challenge of explaining market manipulation within the standard economic Nash framework remains. (For a clue on how to get around this, see the end of this review.)

Tales of manipulation / Even if we don’t seem to know how this works in theory, we still want to know how it works in practice. This brings us to the entertainment portion of the program: Kleit’s always‐​illuminating and frequently revisionist recounting of a number of alleged manipulation episodes.

One example is the (in)famous Hunt brothers’ alleged manipulation of silver in 1979–1980. (See “Silber on Silver,” Summer 2019.) While, as he puts it, “everyone knows” the Hunt brothers manipulated that market, Kleit tells a different story. In the 1979–1980 timeframe, with inflation looming, brothers Nelson, William, and Lamar purchased silver as an inflation hedge. They were so desperate to do so that they bought on margin, ran up against limits set by the Commodity Futures Trading Commission on market positions, and fanned the ire of stakeholders who had sold silver short. However, when Paul Volcker became chair of the Federal Reserve, the likelihood of inflation fell, leading to the Hunts’ bankruptcy. In Kleit’s view, this may have been unwise speculation or hedging, but not manipulation. He points out that while some found that the Hunts’ silver holdings were up to 80% of annual production, the relevant comparison is to the stock of silver, of which the Hunts held no more than 1.5%.

Kleit’s attention to critical facts plays a role in his assessment of the CFTC’s seminal DiPlacido case involving electricity contracts. According to Kleit, the conduct at issue was likely driven only by hedging and the need for liquidity. The allegedly manipulated price is a calculated closing price on an asset—in this case the average price of a particular electricity contract—in the last two minutes of the trading day. The crucial difference between hedging and manipulation is whether the alleged manipulator had sufficient share of these two‐​minute settlements to influence the calculated price and a large enough position of assets priced at, but not influencing, that level.

Kleit reports that what little evidence there was did support hedging, and that the manipulation finding required the CFTC to be “both the prosecutor and the judge.” The CFTC constructed a manipulation case largely by claiming that the alleged manipulator created an artificial price by “violating bids”—that is, not accepting the best offer. Kleit points out that during hectic and volatile times on the trading floor, traders may be more interested in liquidity than getting the best price. Consequently, the CFTC failed to consider alternative explanations and also whether there was motive to manipulate. I would not put much weight on motive—regulation is hard enough without requiring psychiatry—but this lends even more weight to Kleit’s argument for the importance of standards of evidence in distinguishing manipulation from benign trading and careful investigation of whether those standards were met.

A third example involves the existence and role of market manipulation in the implosion of California’s electricity sector in 2000–2001. After the state opened retail electricity markets in 1998, they worked fine for about two and a half years, until a combination of dry weather reducing supply of hydroelectric power and increased demand from Las Vegas raised wholesale electricity prices in 2000. This led to retail price increases that proved politically unpalatable, leading regulators to cap retail prices while letting wholesale prices rise—a recipe for utility bankruptcy and market collapse.

Since opening California’s electricity markets passed the legislature unanimously and was supported by virtually every stakeholder, demand for a scapegoat may have exceeded demand for reliable electricity. Enron, surely a bad actor in other respects, perfectly fit the bill for the needed villain. However, Kleit finds that Enron’s alleged manipulations—through acquiring rights to transmission without intent to actually use it, or short‐​selling electricity—either were done by many others or took advantage of peculiar market designs. Kleit suggests that if anyone manipulated electricity markets during the California electricity crisis, it was the major utilities, which attempted to drive down wholesale prices by bidding below their expected demand.

Kleit discusses a number of other putative manipulation cases involving transmission congestion, natural gas futures, gas and electricity price indexes, baselines for calculating compensation for reducing use, and others. In these cases and others, his explanations are sardonically entertaining. If you find yourself at dinner with him—as I have many times—and he starts to say he’s got a story about market manipulation to tell you, do not feign a call from the babysitter as an excuse to leave. Stick around and you’ll have a good time. Short of that, you’ll have to read this book.

Lessons / Unlike most dinner parties, however, this book lacks a conclusion. The reader is left to extract the lessons from Kleit’s case studies. I took away a few.

One is that, as he repeatedly admits, stories of energy market manipulation are usually confusing. When the underlying theory remains unclear, that’s to be expected. A second lesson is that most alleged manipulations stand a good chance of being either benign hedging and speculation, expensive mistakes, or exploitation of dubious regulatory market designs. Third, in the one or two instances where Kleit thinks manipulation may have taken place, the regulatory proceedings were dubious at best and abusive at worst. If there’s a clear and clean case of manipulation in energy markets, it is not in this book.

So, how to explain the existence of successful manipulation? While manipulating an index and other forms of lying are inconsistent with the Nash equilibrium method for understanding strategic outcomes, it may not be if listeners believe they are being told the truth most of the time. If, say, investors believe indexes are truthful 90% of the time, then they may choose to always believe what they hear—perhaps with some hedging—and one observes manipulation 10% of the time.

Why might investors have such optimistic explanations? One possibility is simple trust. For those with a more cynical view of commercial human nature, the belief has to come from something else. The most apparent “something” is enforcement policy to maintain the veracity of an index. Perhaps paradoxically, the justification for going after market manipulation may be that investors simply expect regulators to do so.