Blog for Trading Success

July 1, 2008 – 12:00 am Cross ref: www.awanginvest.com

As the mantra of the Girl Guides or Boy Scouts go, Be Prepared. http://www.pinetreeweb.com/B-P.htm

Having a contingency plan is a form of trade management, I am re-calling an incident I had last year. There was an electronic failure on GLOBEX, and it affected my exit for ES. However, I was lucky to exit later with just 3 pips off.

This is a good lesson in having contingency plans in place, and while this is still fresh in my mind, I would like to share some good lessons here.

When electronic systems fail: Is your Disaster Plan up to date? With big volatility hitting the market, as we are seeing lately, a key electronic trading execution conduit can shut down.

On July 25, 2007, around 10:25 a.m. EDT, the Globex trading application that matches trades for the Chicago Mercantile Exchange shut down. So anyone trading e-mini futures through that conduit could not place or cancel orders for about 15 to 20 minutes.

This would be a problem for any trader, especially day traders. However, this outage happened in the middle of a 20-point drop in the S&P index! So it went from being a problem to a potential disaster.

The setting was in reaction to a strong down move, as the market was making a relatively weak rally when a double dose of bad news hit: the Chrysler leveraged buy-out and existing home sales were much lower than already reduced estimates, so the market started tanking.

Trying to take profits off the board was not nearly as stressful as trying to get out of a trade that was going against you. But not having any control is stressful in itself. I had my stop in the market and was just moving a profit target order when Globex died. Another good reason why we must have our stops in, as my mentor emphasized many times.

It was time to hit the disaster plan:

1. I disconnected and reconnected my execution platform - order rejected.
2. I picked up the number from inside my phone log to dial direct – line
   busy.
3. With the market still moving in my direction, I could afford the time to
   shut down and restart the whole trading platform – order rejected.
4. I finally got a pop-up message that Globex was down. That was why
   I could not get through to the broker – thousands of others were in the same pickle.
5. So I had to keep hitting the exit button and hope that Globex would come back up ; after an eternity, “order filled”.
6. About 5 minutes later, the system finally gave a pop-up that Globex was back online.

I really learned several useful lessons from this episode.

By having a Disaster Plan in place:

1.It will give me a sense of calm as I can follow the steps than if I were to come up with the steps when disaster strikes. The direct line to our broker should be at hand.
2. It does not mean that I can avoid the consequences of the disaster, but I can take actions to limit the effects of the problem.
3. With the exchange’s communication conduit down, no trader can get through.
4. Because we cannot always avoid disasters, position sizing has shown its importance especially for highly leveraged instruments like futures, forex or options. ALSO our stops should be in the moment the order is filled.

I hope to use these lessons to prepare for the next hiccup that comes along; come it will as the market chops around.

ANA aka IDKIT

Ag Moderator

Bookmark to:

In relation to last night’s post, Manish first asked: “Which volume number should one use?”

I use the average volume contained in a swing. But before I discuss that idea, we need to address the second question: what data source do I use? To answer that question we first need to understand the impact of the quarterly futures roll on volume.

Figure 1 shows CSI’s Perpetual contract for the S&P Futures. Notice how the volume peaks at roll-over and declines and starts the process all over again. This ‘roll-volume’ affects the volume in the cash because the US market is relatively efficient; consequently, there is some relationship between cash and futures volumes during futures roll-over.

FIGURE 1 CSI S&P Perpetual

Figure 2 shows the Cash S&P for the same period as Figure 1.

FIGURE 2 Cash S&P

I prefer to use normalized volume i.e. volume that identifies ‘normal’ volume for a period and then is used as a reference to plot ‘today’s   volume ‘. The best service for the US Stock Markets is Market Volume (http://www.marketvolume.com/). Figure 3 shows Market Volume for the same period as Figures 1 and 2. (Yes, the service includes intra-day volume).

FIGURE 3 Normalized S&P Volume

For other futures markets I use software I had written for me to produce normalized volume.

(My program is not commercially available. However, a STC student, Kym Haines <khaines@bigpond.com>, created a program that takes a data base in ASCII format of normal contracts and converts the contracts into one with ‘normalized’ volume. This is an excellent program. I subjected it to rigorous testing and the results produced were robust.

Kym made if freely available to STCers; I don’t know if he will sell it. If you are interested in buying, you can always ask him).

So for cash indices, I use normalized volume of the share volume transacted at the exchange; for futures contracts I use normalized volume derived from CSI’s data. As we may expect, the futures contracts provide a much greater contrast between normalized volume and that reported by the exchanges.

Now that I have answered that question, let’s turn to ‘average volume’. All I do here is average the volume in the data set I want to measure. Market-Analyst (http://www.market-analyst.com/) has a tool, ‘the probability box’, that does this automatically for me over the range I choose.

Finally as far as Open Interest is concerned, I use it in COT data format. For this I rely on COT services produced by Floyd Upperman (http://www.floydupperman.com/) and Steve Briese (http://www.insidercapital.com/) [The service is called: Bullish Review of Commodity Insiders]. Note that Steve has 2 other advisory services - Insider Futures and Insider Currencies. I don’t use these.

Bookmark to:

I wrote this piece last night but when I awoke this morning I found it has ‘disappeared’. The wonders of modern technology.

I just returned from Vietnam and found it an interesting experience. Last year, it had one of the most buoyant stock markets but at the time of my visit then, cracks were already apearing in the secular uptrend. This year Vietnam is experiencing a bear market. Indeed, when I looked at its chart, I saw the same pattern repeated across the region - the exceptions being Thailand, Taiwan and South Korea: most markets have completed a sharp first leg down; some have made a second leg rally while with the others, we are still awaiting that event.

In the US, I find the stock charts (DJIA, S$P etc) equivocal. Take the S&P, for example.

For the bear market to be confirmed, we need to see the market accept below 1257 in the S&P. However, we do have some benchmarks and clues to lean against before 1257.

1. If we look at the up and down legs since the May 19 high (1440) and compare the price/volume relationships, we find that the range and volume of the up legs has shrunk - 30% and 5% respectively. The reduction of 5% (volume) is at best an indication, I would not like to make too much of it. The reduction in the range is worth noticing so I’d lean to the correction off the . For me, the next down leg will be important, if we see stronger volume come in, I’ll say we had a small leg 1 down and a leg 2 correction.
2. The critical levels to watch will be 1373 and 1440. If the market accepts above 1440, I expect to see the Primary Sell Zone 1576 to 1557 tested. In addition a bull-bar close above 1424 will strongly suggest acceptance above 1440. If the market accepts below 1373, I expect a retest of 1270 to 1257.

(All figures basis cash S&P)

Bookmark to:

Gary asked:

“Hello. Could u advise if the tip and bottom prices of value areas are of more importance than the POC since effectively if we take positions at the previous day POC, we would be subject to whip saws as it rotates ard there?”

The AUDUSD 290-minute chart is a perfect illustration for the answer to this question. Let’s turn to Figure 1.

FIGURE 1 AUDUSD 290-Minutes

In Barros Swing terms, we are waiting for a 5-period line to turn down to complete ‘C’ and confirm a sideways market. But the Market Profile shows a probable bell curve (sideways pattern) where the Value Area is forming . The Boundaries of the Value Area are .9636 to .9582.

Notice the price action at the top and bottom of value. We bounced off .9636 and .9582. Figure 1 was taken this morning Hong Kong time. If you look at a chart at time of writing, you’ll see the market bounced of the two zones about twice during the day. This is the way I’d expect the top and bottom of value to behave.

Now look at the Point Of Control (POC) (the red line), notice how the market trades above and below it. In a congestion market, prices are attracted to a POC like a magnet but the POC itself does not provide support.

On the other hand, where you have a zig-zag or straight line correction, the POC may form support or resistance. In Figure 2, the red line is the POC of an incomplete distribution. Notice that the POC did form support for 5 weeks before the market broke to the 50% retracement of the Initial Price Movement (IPM).

FIGURE 2 GOLD

So, whether or not the POC will provide support or resistance depends on the type of correction.

Gary also asked: “Essentially i am always in a dilemma if i shd enter prices at top of the HVA or low of LVA in anticaption of a breakout from the previous day value area or enter trades at POC of the previous day”.

In reply:

1. I said that I would not enter at the POC of the previous day. You now know that this answer assumed a sideways market in progress.
2. I also said that the first part of the question would depend on context. In determining context, one tool I have given you is the algorithm that gives you the probability that development has completed (http://tradingsuccess.com/blog/end-of-developement-nature-of-trends-alogorithm-364.html).

In Figure 1, development for the Value Area began at ‘C?’. Apply the alogrithm to determine if formation of the Value Area has completed. As long as it is still incomplete, then we can sell off the top and buy off the bottom of Value. But if formation of the Value Area is complete, then I’d trade the breakout.

Hope this helps Gary and the other readers of this blog.

Bookmark to:

I received an email asking for my details about why I exited the ES trade. I’ll make that the subject of tonight’s blog.

In “Acceptance of Decisions’ Consequences” (http://tradingsuccess.com/blog/371-371.html), I said “Over the weekend, I decided to exit the longs mainly because the rally since May 9 had been on declining average volume”. This is the comment I have been asked to elaborate on.

In this case, I performed two types of volume analysis:

1. A Ray Wave average volume analysis and compare the average volume of the previous impulse moves down with the average volume of the current impulse moves up.
2. I also compared the average volume of all the up impulse moves in the current up swing.

Figure 1 shows the volume analysis. I have colour coded the pairs for easier comparison. The rally up started well. The first leg (blue) up was stronger than the first leg down. Each succeeding leg was about equal until the magenta leg when the up leg was a full 12% less (375k:421K) than the corresponding down leg. Given that the market had broken through resistance with this move, this was not a bullish sign. We only needed a strong day down to confirm the bearish signals.

Even more important, on the up legs, each succeeding thrust was on lower volume. This is not a volume picture I assess as healthy.

FIGURE 1 Volume

There was one final tell tale warning - the breakout volume. In early May we took out the highs set in February. The market was unable to hold the highs, and pulled back. Then we had an apparently bullish breakout BUT the breakout was on even lower volume than the earlier May breakout! At this point all amber lights were flashing red.

That’s the volume picture. Let’s turn to the price action of the breakout on May 16.

May 16 was a small range day - not the sort of day I want to see on a breakout. True there was a rejection of the lower extreme but given the volume configuration, it was enough for me to exit all longs in the night session on Sunday Central Time (CT). If I hadn’t exited longs by Monday May 19 pit session open, I’d certainly have exited by end of trading May 19. Figure 2 shows the price action.

FIGURE 2 Price Action

The signs were there for all to see. Yet I have acquaintances who have set stops on longs below 1380 rather than exiting early - ‘it may go up’ is the refrain. If stopped out, the trade will have cost them 51 E-Mini points. I’d much rather take my trade of a small profit than risk a 51-point loss when I see the market flashing amber and red warning lights.

I trust that answers the question posed.

But I want to emphasize that the ES result was one side of the coin. There is another side to this coin and you need to to be able to accept it. My Crude Oil result shows the other side.

I was stopped out of the remaining 10% of Crude Oil positions at 131.975 and now the market is back above 133 and may trend higher.

Now, I tightened my stops and took profits on the majority of the position as the market went to 135 at an average price of around 130.5. I did this on the basis that the market was ripe for a pullback and because May 22’s price action in Crude was in the same category as the ES breakout price action i.e.price action that for me signaled strong warning signs of a strong correction.

If the Crude Oil market moves to 150 without a 13-w retracement, I’ll have left money on the table. Note that I will not re-enter given my assessment that at these levels I believe it is a high risk trade.

But whatever the market does, I accept that I made the best decision possible given the state of my knowledge and personality. It is this acceptance that is the single most important factor to my success - if I were forced to choose one factor.

Bookmark to: