Technically Speaking, June 2012

By now, the story is well known. In early May, JPMorgan announced that some hedges designed to limit risk had not worked as planned and the bank was facing losses of at least $2 billion. The trades were on certain tranches of credit-swap indexes.

The MTA enjoys a diverse and growing membership and some members may be involved in these specialized markets or even been on the winning side of these trades. It wasn’t always this way. Originally, the MTA limited membership to technicians focused on the stock market and living in New York. Founding members quickly expanded the organization and today it is possible that a member is an integral part of any news story. It is also possible that some members barely follow the news, profitably focusing
solely on charts and ignoring anything else.

Several features of the JPMorgan story show how technical analysis has grown and members are using similar trading strategies and techniques, although most members are not risking billions of dollars in trading capital. This story also provides an opportunity to highlight that technical analysis is a valuable tool for risk management.

The instruments involved in the JPMorgan trade are complex financial instruments. They are derivatives based on credit default swaps which are also derivatives. Credit default swaps pay the buyer the amount they lose if a borrower defaults and fails to meet its debt obligations. Credit-swap indexes aggregate a number of credit default swaps into a single instrument using the same underlying idea for indexing that Charles Dow used to create the Dow Jones Industrial Average in the late 1800s.

Tranches break the index down even further and are available with varying degrees of risk. In some ways, tranches slice up the index into pieces. Some slices will have more risk than others and the price of each tranche should be set by the risk of the underlying credit-swap index and the characteristics of the tranche. The exact variables involved with pricing these products are
probably best considered as fundamental analysis and require detailed and specialized knowledge.

Although identifying the specific price that each tranche should trade at can be a complex math problem, they do eventually trade and the price moves are based on supply and demand. This is a principle that the traders at JPMorgan seemed to ignore. According to Bloomberg, “Bloomberg News first reported April 5 that [the trader in London who is at the center of the problem] had built positions that were so large he was driving price moves in the $10 trillion market for credit-swap indexes.”

Most traders will never face this problem, but acquiring too large a position (once known as cornering a market) has led to problems for traders for at least 150 years. On a day known as Black Friday in 1869, two speculators tried to corner the gold market on the New York Gold Exchange. Massive government selling of gold then led to a crash in the price and many speculators lost money. On Silver Thursday in 1980, a corner in the silver market collapsed. There are other examples throughout history.  Technicians understand that market history tends to repeat itself and study incidents like this to learn how to avoid those problems. While most traders never control enough capital to move market prices with their trades, this market reaction is certainly possible in thinly traded stocks or options markets that many members are active in. While the JPMorgan trades are not in the history books yet, they are still providing lessons to traders.

In announcing the troubled trading position, Jamie Dimon, Chief Executive Officer of JPMorgan, noted that the trades were “riskier, more volatile and less effective as an economic hedge than we thought.” He added, “But in hindsight, the new strategy was flawed, complex, poorly reviewed, poorly executed and poorly monitored.”

Traders can draw lessons from each of those problems:

• Flawed: Trading strategies should be developed based on a logical premise that withstands the test of time. There are three basic premises of technical analysis – the markets discount the future; prices move in trends; and history repeats itself. Sound trading strategies should be able to point back to these premises as a source for the underlying logic behind the trades.
• Complex: When the MTA was founded, traders relied on chart patterns and a few indicators. Computers did not add indicators with a click of the mouse and optimization was a manual process that could require days with pencil, paper and a calculator. Simple strategies formed the cornerstone of technical analysis. Over time, complexity has been added but systems which become overly complex are unlikely to work as well in the future as they did in the past if they are based on flawed logic. The more complexity a trader adds, the more likely the strategy is to be curve fit to the past and although the future may be similar to the past, it will not unfold exactly the same way.
• Poorly reviewed: While this probably means something different to everyone, one interpretation of a “poorly reviewed” trading system would be one that is not adequately back tested.
• Poorly executed: Traders need to accept feedback from the markets. Analyzing trades to determine if they are being executed at a low cost is a practice among institutional traders and should merit at least a cursory glance from individual traders. Simple changes like trading at different times of the day could impact execution and trading performance.
• Poorly monitored: Almost every trader reviews their account balances daily. Justifying losses is a problem for some traders who believe “this is just a short-term move and I’m in it for the long-term” or offer other excuses. Successful traders don’t  argue with the market and monitor performance so that they can minimize losses and maximize gains.

Another lesson to draw from this loss is that portfolio risk needs to be managed at all times. JPMorgan and other institutions often rely on a measure called value at risk, or VaR, which quantifies how much the trader believes they can lose on their positions on 95% of trading days. Specific calculation methods vary but in general terms, the following equation can be used:

VaR = ( µ + z * σ) * P
Where VaR = Value at Risk

µ = mean returns
z = left tail risk assuming a normal distribution
σ = standard deviation of returns
P = portfolio value

JPMorgan has said that its VaR for these trades was believed to be about $67 million but a different model showed that the VaR was actually $129 million. As an isolated fact, we can’t draw any conclusions from the amount of VaR. We do know that Dimon said the method used for calculating this metric was “inadequate.” Successful traders use some type of active risk management and limit losses.

In liquid markets, traders can use stops instead of VaR to asses their risk.  They might risk a portion of their capital on each trade. Individual tranches of a credit-swap may not be liquid and stops may not be practical. Individual traders may find that stops are not practical in some equity or options markets. In these cases, individuals often limit the amount of risk they accept by allocating only a small amount of trading capital to the position.  This rule that almost every novice trader learns may have been ignored by
the risk management desk at JPMorgan.

Some MTA members may read about derivatives markets and billion dollar trades by institutions and think that is something they will never be a part of. They can still take lessons from each trading problem large institutions endure, and we can be certain that there will be additional trades that are “flawed, complex, poorly reviewed, poorly executed and poorly monitored” in the future. One of the basic premises of technical analysis tells us that this will happen again since history tends to repeat itself.

Traders interested in learning from the largest trading losses in history can find a list of those losses at http://go.mta.org/315