The Dealer Profitability Index (DPI) is designed to evaluate the performance of the user over different market conditions. The hypothesis is that for any market where there is a movement in the price, there is a potential maximum amount which may be earned (or lost!). If the user could turn the entire position for every given movement in the market, this potential would be realised. We must acknowledge that this is near impossible, as there are spreads and costs to be taken into consideration. So we calculate the potential profit for every price move starting at 5 points up to the maximum move for the session in question and take an average. This results in a figure which is possibly attainable or could even be surpassed. This becomes base 100 for any session. We then take the user's performance in P & L after deal costs and index it against the base. This is a positive number for a profit and negative for a loss. This is also adjusted to the position limit set for the session, i.e. if the user's average position held was double the position limit, the result would have been adjusted by dividing by 2.

The DPI primary use is to index the user's performance and consistency. Obviously a dealer should be able to generate a profit but this is also seen in the light of the risk profile of the institution. Dealers sometimes do have days when losses are made, and this will be reflected in the variance.

Secondly, the DPI can be used to set a benchmark within a dealing room and to raise the level of that benchmark by improving the performance of the weaker dealers. Regardless of the instrument they are trading, the spot environment provides an uncomplicated basis for achieving this.

Thirdly, when banks and authorities adopt the DPI as a standard, dealers may be publicly ranked or dealing rooms can be compared. This also means that the guesswork will be taken out of employing dealers, either by asking for authentication of their index results, or asking them to perform some sessions to validate their score and consistency. The employing bank can set up the sessions according to their trading regime and see if the dealer can adapt to it.

As TraderMetrics operates in an environment close to the market reality, it is extremely difficult, if not impossible, to falsify the results. In its basic set-up, users are restricted to buying or selling on the current market rates. In more advanced sessions, dealers also have to respect the A. C. I. Code of Conduct, which is also available for indexing, as well as performing to specific bank targets.

The DPI will allow the markets to index individual traders and generally improve the level of professionalism and performance in the dealing room as well as allowing management to evaluate perhaps the most difficult aspect of risk management, namely the performance of dealers relative to profit potential.

• DPI Maximum Position
• DPI Average Exposure
• DPI Average RAROC Position Size
• Real-Time DPI

The Dealer Efficiency Ratio (also referred to as DER/hour) is a statistic which compliments the Dealer Profitability Index and is a method of seeing how much the user returns for a given risk. The risk is defined as the amount of exposure that the user has to the market and the basis for this is the average exposure. The return is defined as the rate of profit and loss after deal costs. The time is the duration of the session. It is calculated by dividing the net profit and loss by the duration of the session over the average exposure and the result is represented as points of quoted currency.

For example: The session is EUR/USD. The net profit and loss to date is USD 1,500, the average exposure is EUR 1,500,000 and the session duration is 2 hours. The DER is calculated as:

This means that the dealer has earned 5 points (0.0001 = 1 point in EUR/USD) per EUR 1 million of exposure.

The hypothesis is that an efficient dealer is one which achieves the maximum profit whilst having the least exposure to market risk by holding a position. For example, a dealer who sees an opportunity for a profit will make a foray into the market to take that profit: He/she will only be exposed to risk for a short time. A dealer, on the other hand, who has a position most of the time in the market and takes profits as and when they appear is exposed to market shocks, and may suffer losses because the position cannot be liquidated due to sudden lack of market liquidity.

This goes somewhat against the philosophy of the DPI, but management may see this in two ways; Limits (and therefore capital) is given to a dealer to be applied in the markets and therefore a dealer must always be looking for the maximum return on all of the capital. Or, the alternate view is that a dealer should not over-trade, not have a position for the sake of having a position and protect the capital against market shocks by holding the major proportion in reserve. Other management would prefer to see their dealers jobbing.

In dealer evaluation, both performance methods are considered and it is up to the institution to decide which one has more relevance, but both should always be included in dealer records. The DER is also a component of the RARE measurement which takes volatility into account.

The Risk Adjusted Ratio of Effiency is a metric based on the Dealer Efficiency Ratio but taking into account the volatility during the session. It is calculated by taking the DER and dividing by the volatility. For example, the DER of 5 and a volatility of 13.77% would result in a RARE of 36.3108. The reason that volatility is applied to DER is that it is more difficult for a dealer to earn profits (especially jobbing) in a quiet market, whereas a volatile market should result in a high DER number.

Risk Adjusted Return on Capital (RAROC) was invented by Bankers Trust in the 1970s as a way of adjusting individual dealer's profits for risk. This made it possible to compare different traders in different markets and instruments. This meant that it has become possible to see which dealer performed better and how much capital should be applied to the operation. RAROC adjusts profits for capital at risk, which can be defined as the amount of capital needed to cover 99% of losses annually.

TraderMetrics applies this metric to the intraday dealing operations with meaningful results. The calculation (as used in TraderMetrics) is based on the volatility, the confidence level, the position limit and the net profit & loss.

To take an example for a EUR/USD session: The position limit is set at EUR 10,000,000. The volatility of the session is 13.77%. From this information we can calculate the amount of capital to cover 99% of the losses. 1% of the normal distribution lies 2.33 standard deviations below the mean (our confidence level), the worst loss possible is calculated as:

Note: This number is not displayed in TraderMetrics

We want to estimate the annualised return on capital based on this risk adjusted capital. So we divide the net profit and loss converted to EUR by the capital, EUR 3,208,410. Assuming a theoretical profit/loss of EUR 2,500, the calculation yields:

To annualise this we apply the 'square root of time' function. This function is described in detail elsewhere in this section but for a session of two hours, the calculation is:

This number can be used on a real time basis as the most effective method of monitoring a trader's performance. The trader can us the number as a decision making tool to set stops for the position or use as objectives for profit taking.

For instance, at a certain trading bank, a dealer who loses more than 10% of the RAROC capital in one month can no longer trade. Therefore, a dealer can constantly limit losses to -10% of RAROC Risk-Adjusted.

The Risk Adjusted Return on Turnover is a metric designed to show how efficient a dealer is in earning profits in proportion to the turnover. High turnover leads to high expenses. The higher the turnover or volatility, the lower the RAROT number. Dealers should aim for the highest possible result. To calculate RAROT, we divide the net profit & loss by the turnover (both in quoted currency terms). The result is divided by the volatility and then annualised. This is achieved by applying the 'square root of time' function, which is the square root of the session duration as a fraction of a year multiplied by the factor.

For example, the Net Profit & Loss is USD 1,500, the turnover is USD 25,879,000 and the volatility is 13.77%. The session has lasted 2 hours. The RAROT is calculated as follows:

Profit Targets and Loss Limits are designed to discipline the dealer into attaining a set level of profitability which takes into account the dealing room budget, the position limits set and the competence of the user. On the other hand, limits on losses are designed to limit the dealing room's exposure to trading losses. It is of greatest importance that a dealer learns to adhere to these limits. The management of a trading institution expect a return on the capital employed and a return on their investment in salaries, equipment and communications. Every manager must have a target, which should be reasonably attainable given the market conditions, at the same time knowing how much loss, if any, can be tolerated. These limits should be dynamic on a day-to-day basis and be based on budgets, profit or loss to date and market conditions.

The dealer should not go beyond these limits. Of course if a windfall profit is possible, then it is worth holding the position until the maximum profit is realised, but this should be done with the involvement of the chief dealer. In the position statistics, TraderMetrics shows if and by how much these limits have been broken. It is quite an achievement to be 1 million down and then bring the loss down to 10,000, but the management must look on this as a potential loss of 1 million, because if the trading had to stop at the time of maximum loss, this would have been realised. Conversely, if their was a profit of 1 million that had not been realised and had been reduced to 10,000, this could be considered a loss. (Opportunity loss).

So the rules are: On attaining the profit target, cut back or stop dealing activity. On reaching the loss limit: cut the position and stop dealing!

Profit Take and Cut Loss settings are designed to address perhaps the most fundamental problem faced by a dealerindeed any investor: When to take profit and when to cut a loss. Despite the principles seeming obvious, many dealers run their losses and cut their losses, instead of the opposite, due to fear, lack of knowledge of their own psyche and lack of self-discipline. Using this tool with any dealer will help to overcome these problems. The system works by the trainer setting a ratio of losses to profits. Obviously the profit aspect should be higher than the loss. When the session starts, the user will endeavour to minimise the losses in points of any position to the target value and maximise the profit by attaining the profit target. Although this works for any session, the best way to use the system is to set up a session where incoming calls, outgoing calls and electronic broker are turned off and the maximum and minimum deal limits are set to the same value, say 5 million.

The dealer then has to take a position for 5 million and try to reach the profit target and on attaining this, cut the position. If the position turns loss making, the position is cut before the loss target is reached. When this is done several times during a session, the statistics may be observed and the overall performance noted. If a new dealer is taught on this system, he or she will learn the self discipline and also be able to master the fear that many experienced dealers have when faced with a loss.

The optimum ratio, in our opinion, is 2.6:1. That is, the profit take or limit 'order' should be set to 2.6 times the stop loss. For example, 52 points limit against 20 points loss. This ratio is derived from experience. Too low, (under 2) and profits in the long term will not be achieved. To high, and the probability of such a high ratio diminishes. Think of odds on horse race betting.

This training should be combined with monitoring the ratio of winners to losers. Any system will rarely produce more than 50% winners. Having a 2.6:1 Risk Return Ratio (RRR) will allow you to still have less than a 50% win/loss ratio and make money, because it generates a positive mathematical expectancy (PME).

Mastering these concepts, together with money management, will lead to consistent profits, no matter what asset class you are trading