16 Daylrading

A day trade is a position entered and liquidated during a single trading day. The techniques of time of day and daily,patterns, discussed in such ngid form in Chapter 15, are put to use by the dsy trader. Day, trading requires extreme discipline, excellent planning, anticipation, and concentration. The need for a fest response to changing situations tends to exaggerate any bad trading habits as in other fields, the shorter the response time, the greater the chance for error. In this chapter, we will extend the idea of clas- trading to sjstems and methods that may also be held ovemight, but expect to limit the trade to about 24 hours.

To keep mistakes to a minimum, each days strategj must be planned in advance. It should focus on the mosl likely situations that might occur based on the nature of the current price movement. There should also be a contingency plan for the extreme unexpected moves in either direction. Making spot decisions during market hours m-ill cause more frequent errors.

Computers have caused the number of dsy traders who rely on sjstems to increase, and have created an entire class of screen traders. The steadj- increase in automated exchanges, led by Germanys DTB and followed by Frances conversion of Matif. is sending a strong message that electronic trading will dominate the future. The availability, of intraday price feeds and sjstem development platforms, such as Omegas TradeStation. have greatly increased the number of dsy-trading participants. Only, 10 years ago the best one could expect was to display standard indicators and moving averages on a real-time price chart. Now you can fully program complex trading strategies that combine more than one market and more than one time fiame into a single padage and display bus, and sell signals on the screen as well as record a historic log of trades. All of these tools have greatly increased competition among individual and commercial traders.

For the arbitrageur, computers have had an even stronger impact. Sophisticated sjstems at banks and large financial institutions consolidate data feeds that bring current transactions on every tjpe of interest rate vehicle in even-maturity and major currency. Analytic programs can find issues that are outliers and show which combinations (called ships) can produce a riskless profit. For the individual trader, few of these opportunities are available, although they add liquidity to the marfcet. Individuals, however, find it much easier to create spreads of different deliveries -„,ithin the same marfcet as as spread between two related products. Stock traders can create sector badets or loofc for performance differences within a sector. Spread tradirig and formulated values, such as the energj cracfc or soybean crush, can be improved using similar displajs. Many of the opportunities that now seem so easj to see would previously, have been missed. This faster, more sjstematic response to the marfcet allows traders to improve profits and reduce rid; in anydsy-trading method.

IMPACT OF TRANSACTION COSTS

Transaction costs are the greatest dderrent to dsy-trading profits. The failure to execute near the intended price, plus the commission costs, can remove a large part of potential

profitability and even tum expected profits into losses. An aspiring dsy trader has two wajs of improving performance: by pajing lower commissions and by carefiil selection of opportunities. Table I6-I presents the percentage relationship of moderate transaction costs to the maximum daily price move. The average dollar volatility for the years 1990 through 1996 is shown next to the percentage represented by a $100 transaction cost. In general, $ 100 is a modesl value for the combination of commissions and slippage, and traders would be very pleased to extract an average profit of one-half the daily volatility. A practical approach would therefore use twice the percentage effect of a $100 cost shown here.

Realistically, a dsy trader would choose the index marfcets, more volatile currencies, and long-term interesl rates as their best opportunity for profits. Com represents the least desirable marfcet because very low volatility leaves more rid; than opportunity. Although a $ 100 transaction cost may seem high for this marfcet, a $20 commission and slippage equal to a minimum move of I/4c, or $12 JO, totaling $45, would be the smallest possible costs. An occasional fest move in the marfcet, based on unexpected economic news, or large orders entering at one time, must result in a larger number, raising the average.

To fceep dollar volatility of a futures contract to a level 1 1 by most traders, exchanges have been forced to resize contracts, specifically stock index marfcets. as overall share values have climbed. During 1997, the S&P was reduced from $500 per basis point to $250, and the FTSE-IOO fian £25 to £10. In 1998, the French CAC40 conhaa was cut to 25" of its previous size. This increases the relative size of transaction costs

Liquidity

The importance of hquidity is magmfied in day trading. Execution slippage of $ 100, measured as the difference between the sjstem price, or Market Order price, and the actual filled price, will have litde impact on a monthlong h-ade netting $2,000; however, it will be critical for a dsy trade with a profit objective of under $300 The selection of dsy-trading candidates begins with those maikets of greater volume. Whether one contract or 1,000, a

TABLE 16-1 Volatility and Liiuidicy, U.5. Maritets. 1990-1996

thinly traded market produces slippage that will cut 8 1 into profits. In choosing among index maikets that usually offer the greatest volatility and profit potential, you are alwajs safest with the markets that combine the highest volume and highest volatility

Occasionally, markets with light volume show larger price moves than similar markets traded on other exchanges. Traders will be tempted to profit from these moves but A-di consistently find that the execution of an order at the posted price is elusive. The reality of trading these maikets is that a Maiket Order is not advisable due to the thinness of the trading, a Limit Order may not get filled, and a spread is not quoted at anything resembling the appareni price relationships that you see on a quote screen; there is no real way to take advantage of these perceived profits. If an execution succeeds, exiting the position still has the same problems plus some added uigency

Missed Orders

Because the profit objective, projected daily range, holding time, and end-of-dsy constraints all put limits on a dsy trade, there are situations in which the market jumps after anews release and you cannot get filled anjwhere close to your intended price. With most of the potential profit gone before you enter the order, it would not be surprising to simply skip that trade. These missed orders, called unables, can add up to a large part of your profits; at the same time they never reduce your losses.

Some markets are prone to more unables. For the energj complex, heated militarj or political activity in the Mid-East can cause a prolonged period of very erratic price movement, resulting in as much as 20°o unexecuted trades during a 1-month period. If we consider the normal profile of a short-term frading system as having an average net profit of $250, an average loss of $150, and a 50°o frequency of profits, we expect a profit of 85.000 for every 100 trades and a reasonable profit-to-loss ratio of 1.66. If, however, there are 10°o unables, that missed opportunity must come from the profits; if the market was moving in the opposite way, you would get all of your positions filled. Then 10°o of the profitable frades means 5 frades out of every 100 for a total of $1,250 missed. This reduces the total profits to $4,750 and the profit feet or to 150. It may tum a marginal frading sU-ategj. or one with small profits per trade, from a profit to a loss

Price Ranges

The average daily range (not the true range) shows the maximum potential for a single day-trading profit. Some markets, such as gold and com, combine a lack of liquidity with a narrow daily range and are easily disqualified from a day trade opportunity. Eurodollars have extremely high volume but very little movement; therefore, they are also not a good candidate for day trading. When volume is combined with volatility, the world index marfcets, followed by the currencies and long-term interest rates, are shown to he the best choices. Eneigj marfcets represent the next tier, grains remain active, and other marfcets have far lower volume.

Day traders may find that those marfcets with hditional daily trading limits present a problem during high-volatility periods. Day trading does best in marfcets that have wide swings not deterred by exchange limits; a single locked-limit move can generate a loss that offsets many profitable day trades. High volatility and locked-limit moves present a contradiction for day trading. Rules that provide for expanding limits have greatly helped reduce the frequency of locked-limit dajs; however, traders must ahvajs be on guard for this situation.

Estimating Slippage Costs

If you know the cost of slippage, you can do a much better job selecting the marfcets to trade and have a realiatic appraisal of your trading expectations. The factors that make up

slippage are volatility, overall market volume, current marfcet activity, and the size of the order being placed. Of these items, current marfcet activity is the most difficult to record, because it requires some estimation of volume as it accumulates throughout the day. While actual volume is not available for most marfcets, a reasonable approximation can be made using tick volume, the number of price changes during a fixed interval. In general, tick volume is directly proportional to actual trading volume; during periods of greater activity, both confract or share volume will increase along with the number of price changes, if you have carefully recorded the order price, execution price, volatility (daily high low), daily volume, time of day, and tick volume, you can find the importance of each factor and estimate the slippage for any trade by appljing the model:

Actual slippage ai time + e, x volanlity + «i x daily volume +

e, X current activity at time + , vsize of order

By creating a spreadsheet of values for all trades, you can solve for ao, ,, az, «3, and a. and then estimate future slippage at any time during the day, given the current volume and order size. This approach was taken in 1992 for large Stop orders using a ratio of order size to current market activity. The results, shown in Table 16-2. show that slippage was very modest for most markets. The far right column gives the slippage likely at the 5% level; that is, there Is onXy a 5% chance of having slippage greater than S135 per contract for sugar, combining both entry and exit. Commissions were not included The unusually large "worst" loss in Treasury bills was confirmed as correct.

Price Level

Continuing volatilfty Ife most common in marfcets that are at abnormally high price levels. Bull marfcets are followed by bear marfcets, and the combination creates sustained activity Prices that are volatile at low levels are mosl likely to be at the end of a decline and can be returning to previous, less volatile pattems. The agricultural marfcets. when carrjover

Th.jnas V ir. WleBrorsen. and £b-Miin Liu. " e Costs ..f a L.Me Technical Tra.ler." Tecmical Analysis ..f .- & Cc . . .l, :.

TABLE 16-2 Slippage (Loss) for Stop Orders (Slippage in Dollars per Confract)