On Thursday, May 5, commodities markets fell sharply. On that day alone, the DJ-UBS index of commodities prices lost 4.7 per cent. The rout is said to have started in the market for silver, but rapidly spread to a wide range of commodities.
Silver is used for commercial purposes. Recently, however, the metal has been trading predominantly as a speculative asset. A lot of people now associate silver with the role traditionally given to gold as a store of value against the erosion of fiat money. As such, the price of silver, instead of reflecting the value of the stream of income derived from the traditional commercial uses, is driven by changing beliefs about grand macroeconomic events. This is how silver has gone up by more than 70 per cent in just three months, between the end of January and April 29.
Silver prices halted their rapid ascent with an abrupt fall of 28 per cent in the first week of May. Fundamentals can hardly explain such a violent reversal. Searching for an answer, many pointed their fingers at the Exchange, which raised margin requirements on futures positions four times for a total increase of 61 per cent since April 25. Three increases happened in just five trading days. The argument goes like this: many investors, surprised by the repeated large margin increases were unable to quickly arrange financing of the margin and were forced to sell their positions, driving the price down. This selloff may have been aggravated as other investors panicked. Some complain that the Exchange was deliberately, and unfairly, trying to push prices down. Others use the occasion to press the political case in favor of more aggressive use of margin requirements as well as position limits to stop speculators from inflating the price in the first place.
The events of the last couple of weeks in the silver market are a useful opportunity to briefly review the role of margin requirements and what is known about their impact on trading and prices.
First, let’s be clear about why a futures exchange requires investors to post margin. It is because the exchange, in clearing the trade, inserts itself into the transaction, assuring each side that it will be paid its profits. To do this, the exchange needs to be sure each side will also pay up on its losses. That is what the margin is there for. The potential for losses is determined by the volatility of the price of the contract, and this volatility varies through time. When the volatility changes, the appropriate margin changes, too. In this most recent episode, Kimberley Taylor, president of CME Clearing, had to come out in defense of the CME’s actions, explaining that the margin increases were a response to increased price volatility and designed to protect the exchange from the risk of non-performance on investors’ positions. She was exactly right to do so. The process looks more subjective than it is because the exchange has lagged in the degree to which it automates the process. Unfortunately, this invites conjecture and debate about other motives for adjusting the minimum margin.
The margins that are generally discussed in the media are the minimum margins that the exchange requires. Individual brokers are free to require larger margins from this or that customer or for this or that contract, and they do. Parties to OTC trades, while free of any externally mandated margin requirement, can include a margin requirement as a part of the transaction, and many do.
The questions this case raises include…
- How do margin requirements affect trading? Are speculators affected more than end-users?
- Do changes in margin requirements affect prices? In particular, how do they impact volatility? Can they trigger market crashes?
- Can OTC markets, which do not require margins, avoid the alleged cause of turmoil created by the margin requirements in Exchanges?
1. How do margin requirements affect trading? Are speculators affected more than end-users?
Every futures trade exposes the trader to the potential for loss as well as gain. To the extent that the loss may exceed posted margin, the trader is betting with money borrowed from his or her counterparty. If the margin level is increased, the trader is losing the benefit of that source of credit. Of course, the trader can go out and replace that lost creditor by borrowing money elsewhere, off the exchange. Changing margin requirements only affect the trader insofar as the exchange or counterparty is the marginal source of the trader’s credit. Someone determined to speculate on the price of silver or a company deliberately hedging the price of silver can maintain their position in the face of a margin increase by covering the margin with money borrowed elsewhere. Only if the cost of the alternative source of credit is too high will the trader adjust its position as a result of a change in the level of the margin required.
The academic literature does appear to find a small impact of margin changes on open interest, so it does appear that margin changes impact some traders. Exactly who and how much is less clear. The impact does not appear to be large. We can look to the data for this silver episode and see similarly modest impacts.
Table 1 shows that as the cost of the margin, measured as a percentage of the value of the contract (3rd column from the right), kept going up, the amount of open interest (1st column from the right), which measures the number of contracts outstanding, declined. From April 29 to May 5, volume fell from 80,420 contracts to 77,140 contracts. Closer examination reveals that on the first three days of the week (May 2 to May 4) open interest declined and the price fell, indicating a predominance of long positions being closed. On Thursday, May 5, when commodities markets felt a big tremor, open interest increased and silver price of silver fell, indicating that traders jumped in short. Finally, on May 9, open interest declined further and the price rebounded, indicating short positions being covered.
Table 2 reports data from the CFTC’s Commitments of Traders report across all contracts for silver, disclosed on Fridays for the previous Tuesdays. In advance of the big price drop, between Tuesday April 26 and Tuesday May 3, Non-commercial traders (speculators) reduced their long positions and increased their short positions, leaving the total number of positions approximately the same. In contrast, Commercials (end users) cut both long and short positions for a combined reduction of 5.5 per cent. Judging from Table 2, it appears that Commercials cut down their positions by more than Non-Commercials between April 26 and May 3, as margins increased 38 per cent. The week after, which includes Thursday, May 5, the day of the steep decline, tells a very different story. Between May 3 and May 10, as margins increased a further 33 per cent, Non-commercials reduced their long and short positions by 18 per cent and 58 per cent, respectively; Commercials, on the other hand, increased their long positions by 6 per cent and short positions by 3 per cent. During the period April 26-May 10 margins increased a total of 83 percent, and Non-commercials reduced their positions (long and short) by 27 per cent, while Commercials reduced their combined positions by just 1.5 per cent. It appears that a sharp increase in margins in a short period of time had no visible impact on end users and a meaningful impact on some speculators.
2. Do changes in margin requirements affect prices and in particular do they contribute to magnify volatility and trigger market crashes?
The literature gives ambiguous results about the impact of margin changes on volatility and prices. Some affects have sometimes been found, but the size is not impressive. A number of studies [Hardouvelis (1988), (1990), Hardouvelis and Peristinani (1989),(1992)] find that margin requirements reduce price volatility. Other studies [Ferris and Chance (1988), Kupiec (1989), Schwert (1989), Hsieh and Miller (1990)] find either no relationship or a weak positive relationship between margin requirements and market volatility. Hardouvelis and Theodossiou (2002) find that, following large declines in prices, higher margin requirements for long positions increase market volatility; and, following large increases in prices, higher long margin requirements reduce market volatility.
To see how the margin changes in silver this time correlate with movements in the level of prices we can look back at Table 1. The light blue rows indicate the days when the CME increased margin requirements in the July Silver Futures contract (between April 26 and May 9). In three of these days silver prices fell, and in the other two silver prices moved up. Critics of the exchange claim that the dramatic fall in silver prices between Tuesday, May 3, and Thursday, May 5, was caused not by any single margin hike, but by the unexpected cumulative size of the change. If so, why did prices recovered after a further margin increase on May 9?
3. Can OTC markets, which do not require margins, avoid the alleged cause of turmoil created by the margin requirements in Exchanges?
The idea that OTC markets have some special way of dealing with stressful situations, such as rapid price run-ups and increased volatility, is nonsense. It would mean that OTC market makers ignore larger expected losses from counterparties and fail to adjust exposures. Consider a trader entering a long position at the close of May 2, when the price of silver was 4608.4. By the end of the next day the trader would have lost $17,475, after the price fell 349.5 (times 50 ticks). Note that the (exchange) margin on May 2 was $14,513, leaving an uncovered debit of $2962. This is why exchanges revise their margins, and also why brokers often ask for additional collateral to top the margins required by the exchanges. Why would OTC market makers act differently and not demand additional assurance?
Perhaps the reason why OTC markets appear to be tranquil on the surface is because these markets are open to professionals and occasional disturbances can be concealed behind the opacity of the OTC market. But not seeing what goes on in private markets does not make them necessarily safer. It simply means that most problems are dealt from within, and when the veil of secrecy is torn, oh boy, it is because the problems have the characteristics of what might be a perfect storm with enormous destruction power.
 Thanks to Yajun Wang, from the University of Washington St Louis for helping with this information.