Sunday, February 18, 2007

Dealing With Price Spikes: A Best Practice in Trading

Note: This best practice post comes from AnaTrader, a relatively new trader with a background in the arts. Over the time I've corresponded with her, I've found her to be not only a Renaissance woman, but a dedicated student of the markets. AnaTrader relates an incident regarding market spikes due to either faulty ticks or manipulated trades. When you're working orders in the book, such spikes can create unwanted fills. AnaTrader passes along a message from her mentor, Ray Barros, who reassures Forex traders about spikes they might see in their charts.

Over the weekend, some market makers' trading platforms Down Under showed a big spike in GBP/USD, a one time high of 300 pips from normal spread of bid and ask prices as per chart which I extracted this Monday morning.

Many have shorted this instrument, including me, with stops in place for the weekend pending the G7 weekend meeting announcements.

I have to say that when I checked my trading platform, I was not hit or stopped out.

I was perplexed and this is what I learned from my mentor:


A spike can be a false price that did not actually
take place in trading, It usually occurs on weekends
when M-East traders play with the prices. Most
good brokers don't carry the data.
Even if they did, it would not be considered
a trade - in trading parlance the prices would be 'outed' .

Am I relieved that there is a trading practice that such a big spike would be outed or disregarded and not carried on trading platforms of renowned brokers.

We could be slaughtered like pigs, all right.

From Brett: The best practice takeaway from AnaTrader's post is that, in the Forex market, it is important to not overreact to false price spikes that may appear. It sounds as though this is just what the traders are trying to cause in the first place.

A different kind of spike is caused by "fat finger" episodes in which a trader might mistakenly buy 10,000 contracts at the market rather than 1000. The resulting purchase takes out quite a few price levels and can lead to further short covering before anyone figures out that this was simply a mistake. A trader who might be working an order to sell above the market would find themselves short and immediately down a couple of points or so on their position. This is not common, but it is an occupational hazard of working orders in the book. Another occupational hazard is working orders in the book prior to the release of an economic report. If the report does not fall within expectations, a spike of price movement created by automated trades can fill you at very bad prices. Unlike the situations mentioned by AnaTrader, these trades will not be "outed". Rather, you'll be out some capital! My own practice is not to work orders far from the market. If I want in at a good price, I'll try to buy at the current bid or sell at the current offer. But I'm not a scalper and, if I'm telling myself I have to be first in the order queue to get filled and make my trade work out, that suggests my trade idea is not so robust.

Bottom line: You have to know *your* market. The meaning of a rogue price spike in one market could be very different from that in another.

AnaTrader Adds:

For your added information and illustration, just before any major economic news release like last Friday's PPI, one could put in a buy stop and sell stop order for say, GBP/USD, to catch the entry if it spikes up or down. If the figures were to be outside expectations, it would spike either up or down so fast that you cannot get filled at your price level unless you place a straddle order of buy stop and sell stop. Once either order gets hit, one can delete the other not filled or leave it as a protective stop. One has to determine in advance which way an instrument will go if the news were to be outside expectations.

For last Friday, I expected the GBPUSD to spike up should the expectations were outside, and only placed a buy stop order just a few minutes before the economic release at 8.30ET.

Quote of economic releases:

The latest read on inflation at the wholesale level also hit the wires at 8:30 ET. Total PPI and core PPI matched economists' forecasts; but since the data won't alter inflation expectations, the report failed to provide investors with overwhelming evidence that pricing pressures are abating. The focus now turns to next week's more closely-watched CPI report.

Since the reports were within expectations, there was no spike in the GBPUSD immediately after release and I just deleted my buy stop order.

For a novice, it is generally not advisable to trade before any major economic release, but in my case, I am a mentor student and therefore, have an added advantage of being coached and guided in this complex trading method, which has given me some experience in such volatile trade entries.

From Brett: This is indeed an advanced trading method. To place orders in the book ahead of an economic release, you need to have a keen sense for how far a market is likely to move under various scenarios. This means studying past releases and reactions to those. It is not at all uncommon for the market's first, knee-jerk price spike after, say, the release of Fed minutes to not be the direction the market ultimately takes for its subsequent price direction. If you can enter at a good price, it's possible to benefit from the rapid reversal. Prudent position sizing is needed, however, to guard against catastrophic losses should the spike become an outright trend in itself. Thanks to AnaTrader and Ray Barros for the excellent perspectives on price spikes.