London, 29 March 2011: Reuters

Profit margins in the commodities business are shrinking as the number of banks, trading companies and investors in the sector expands faster than the underlying hedging by producers and consumers. Profits from commodity trading are notoriously hard to measure. The big banks tend to include commodity profits in a broad category that includes fixed income and currency trading (FICC). Trading houses are mostly privately owned and do not published detailed accounts. Even for public companies, trading profits are almost never broken out consistently from other activities.

Parts of the commodity business remain phenomenally lucrative. But there are signs margins overall are under pressure as the commodity trading space becomes increasingly crowded and firms struggle to find the right mix of clients.

Trading revenues from commodity cash and derivative positions at nationally chartered commercial banks in the United States fell to $618 million in 2010, according to the Office of the Comptroller of the Currency (OCC). Chart 1.

Revenues were less than half the level posted in 2009 ($1.46 billion) and 2008 ($1.54 billion). U.S. banks did better than in 2007 ($295 million), but it was still one of the worst performances in recent years (Chart 1). Trading revenues for bank holding companies (which includes the non-insured businesses of some of the bigger holding companies that account for the lion's share of commodity trading) were $5.49 billion in 2010, down by almost half from $11.00 billion in 2009.

Revenues have declined even as the notional size of commercial banks' derivatives books has continued to increase (Charts 2-3).

AUTOMATED TRADING

The OCC data is consistent with other indications that all is not well in the commodity trading business. CME and ICE have strong growth in commodity futures and options turnover. But rising turnover may be a misleading indicator of how much commodity markets are really growing.

As much as 30 percent of all trades executed in Q4 2010 in energy and crude oil and other leading commodity contracts on NYMEX were done by automated or algorithmic trading programmes. Some trades were conventional deals broken up into smaller automated lots to conceal them and secure better liquidity ("icebergs"). But a large proportion were pure algo or high-frequency trades, which add volume and (arguably) liquidity but not much sustained open interest.

If the reported growth in turnover is adjusted to exclude some or all of the automated trading, volume growth appears less impressive.

In a recent study of equity markets commissioned by the U.S. Securities and Exchange Commission, Boston Consulting Group found high-frequency trading "is seen as a major driver of the massive rise in trading volume over the last decade: from 2005 to 2009, the volume in NYSE-listed stocks rose by three times; during the same period, the number of trades executed each day rose by [roughly] eight times."

One result is the average number of shares traded per transaction had fallen by more than 60 percent in four years: www.sec.gov/news/studies/2011/967study.pdf

High-frequency and other algorithmic trading still accounts for a much smaller share of total transactions in commodity markets but has been growing fast as the trading technology in commodities catches up rapidly with that commonly deployed elsewhere.

HFT and algo trading depends on and promotes very thin profit margins on each trade. It is far from clear that rising volumes on commodity exchanges imply equivalent gains in revenues for exchanges and the leading brokers and market-makers. It is more likely that increased competition is whittling away profit margins for all dealers.

PROP TRADING FALLS

Banks also appear to be scaling back the amount of risk they keep on their books as proprietary traders and market- makers, opting to focus on flow business that matches buy and sell orders back-to-back or at least very quickly.

Goldman Sachs, probably the pre-eminent commodity trading bank, has cut the amount of commodity risk in its trading book in each of the last four quarters. Value-at-risk (VaR) attributable to commodities fell by more than half from $49 million in Q1 2010 to $23 million by Q4. It is part of a broader reduction in risk-taking that has seen total risk across Goldman's trading book as measured by net VaR shrink from $245 million in the three months ended June 2009 to just $120 million in the three months ended December 2010.

Reductions in commodity VaR as part of a firm-wide reduction in risk-taking are evident at Morgan Stanley and other big dealing banks. Banks have started to scale back prop trading and some of the more aggressive forms of market-making ahead of the introduction of the formal Volcker Rule ban. Goldman Sachs, JP Morgan and other banks have announced the closure of several dedicated prop desks over the past year.

Even before the Volcker ban comes into force, regulators have been taking a tougher stance. Conversion or merger with bank holding companies means the former big securities houses now have on-site federal bank examiners embedded into their daily operation, which is likely to result in more intensive regulation and detailed questioning of how risk is measured and whether it is appropriate.

Even without prodding by regulators, banks may be coming to the conclusion that proprietary risk-taking is not as profitable as it once was in commodities and other asset classes. Spreads, commissions and option premiums have all come under more pressure. Directional trades have proved tricky, and storage plays are less profitable as the market gets crowded.

More generally, once the cost of carrying large positions and the attendant liquidity risks to the parent bank are better understood, returns on prop trading look less attractive.

MORE HEDGERS WANTED

The big problem with derivatives as an asset class is that the returns are zero sum. For every hedge fund or bank that reports a trading profit, another institution must have made a loss. For investors, banks and dealers to be able to make an expected profit on average, there must be a sufficient supply of hedgers willing to incur an expected loss on average.

The problem for all commodity businesses and investors is that the supply of hedgers has grown much more slowly than the number of banks and investors who want a share of the action. Expected profits are being shared out ever more thinly.

Talk of a "super-cycle" has made producers wary about hedging in case they forego too many expected gains. Consumers remain reluctant to hedge their requirements more than 12-24 months ahead. Producers' reluctance to hedge is especially problematic, since they are the natural suppliers of short positions. Lack of producer forward selling has become the key constraint on the profitable expansion of the market.

If the commodity markets are an eco-system, hedgers are at the bottom of the food chain. It is their activity that supports everyone else. In the meantime, the market shows signs of excess market making and trading capacity, which will continue to pressure returns for the major dealers.

Ends --

By John Kemp, Reuters market analyst – for Commodities Now.

The views expressed here are his own.