Minutes of chat session held on Saturday, 16-12-2006
Hello friends,
As scheduled, the chat session was held on Saturday, 16-12-2006 at 9:30pm. Only Murthy, rpc and I were present. Initially we started in the main chat room but other people also came into the same room and started discussing various other matters. So we opened a separate room and continued the discussions there. During this transit the transcript of the talks in the main room was lost and only the discussions in the other room could be recorded. I am therefore reproducing the discussions in the earlier period from memory. If any points are missing, rpc or Murthy may add them.
The discussions were mainly on the points listed by Murthy in his post. As most of the other members were not present it was decided to take decisions only on certain points which would not have any effect on the strategies being followed. Other issues which require approval of other members were just discussed and final decisions would be taken after all other members give their opinion, may be in next Saturday's session.
1) The first point was regarding the trading prices. Murthy had already suggested that only Open price be taken which I also had supported. With a little more deliberations with rpc we finally decided that opening price will be quoted in all the trades reported. The main advantages of this are (i) Comparison of different systems would be on equal terms (ii) Any back testing can be easily performed with EOD data and (iii) one need not watch live prices.
THEREFORE, HENCEFORTH ALL MEMBERS WILL REPORT TRADES WITH OPENING PRICES ONLY.
2) The next point was the number of stocks in portfolio. Murthy and I had already suggested the number to be increased to 10. With detailed discussions with rpc and considering the amount to be invested, finally we agreed to increase the maximum number to 15 stocks, subject to a minimum of 5 stocks. In addition the maximum stocks to be bought on any day should be restricted to 2 only. This allows us to trade upto 8 days.
This brought up another point, that of maximum amount to be invested in a single stock. Murthy had suggested a proportional allocation based on available cash balance. But rpc and I suggested this formula should be revised in such a way that the trader should have some flexibility also. So after discussing at length it was finally decided that a minimum of 50,000 has to be invested in any one stock but not more than 2,00,000.
THEREFORE, HENCEFORTH MEMBERS CAN INVEST IN UPTO 15 STOCKS BUT NOT LESS THAN 5. THE AMOUNT INVESTED IN ANY ONE STOCK HAS TO BE AT LEAST 50,000 BUT NOT MORE THAN 2,00,000. THE NUMBER OF STOCKS TAHT CAN BE PURCHASED ON ANY SINGLE DAY IS RESTRICTED TO 2. IT IS NOT NECESSARY TO BUY EVERY DAY. HOWEVER, THERE IS NO LIMIT ON SELLING. ALL SELL SIGNALS SHOULD BE STRICTLY EXECUTED.
3) This limitation on stocks bought on a day brought another point by rpc. If more BUY signals are generated then which two are to be taken? This again relates to the point Murthy has suggested. It is ranking of the signals. The various options discussed included use of indicators such as MACD, RSI etc as well as volume. The discussions were quite lengthy and it was felt that incorporating indicators would require change in code and this topic of making changes to the existing code has to be discussed separately. It was felt that suggesting an indicator requires thorough discussion by other knowledgeable members. Finally it was decided that only change in Volume should be considered. Here Murthy pointed out that volume change should not be a sudden spurt but a gradual increase (as Karthik advocates). For this purpose rpc suggested considering MA of Volume instead of raw volume. I suggested taking ROC of this MA as a measure of rising volumes. The period to be used was decided as 5 days. At the end we agreed to it but also felt that this should be left to the implementing individuals.
AT PRESENT THE MA OF VOLUME AND ITS ROC, THE PERIOD BEING 5 DAYS, IS SUGGESTED AS A GUIDELINE TO RANK THE SIGNALS. MEMEBERS ARE FREE TO USE IT OR NOT TO USE. ANY OTHER SYSTEM OF RANKING IS NOT ALLOWED TILL A THOROUGH DISCUSSION WITH ALL MEMBERS.
4) The next point was the hot topic at present, which is STOP LOSS. After a prolonged discussion on this issue it was agreed that the present method definitely needs a re-look. But it was felt that we three can not make any suggestions on this without all other members joining the discussion. So the present system of Stop loss will continue. Next Saturday it should be possible for all the members to participate and STOP LOSS should be an important issue to be discussed.
THEREFORE THE PRESENT SYSTEM OF APPLYING STOPLOSS WILL CONTINUE FOR ONE MORE WEEK.
5) The next point discussed was that of a common reporting format. Murthy and I had already written about this. After hearing views of rpc also it was decided that we must report in a common format. ALL signals generated should be reported, not only the traded ones. The BUY and SELL signals should be grouped separately, arranged alphabetically and typed in a single column. The trades should be separately listed, all buys and sells grouped separately and ordered alphabetically. In case of BUY, the Stoploss price (which is at present 95% of buy price) should be reported. Rpc will post a model format for reporting.
ALL MEMBERS ARE REQUESTED TO REPORT IN THE MODEL FORMAT.
6) An additional point was brought up by me. It was regarding changes to the system. I have discovered a shortcoming in my present system where some BUY/SELL signals do not get generated. The correction is available but not included so far. So, I asked whether I can make the necessary changes. Murthy agreed that creator of the system should have the opportunity to change it. He also said all the creators of the systems should also get the chance to change. Rpc also voiced that he also is not completely happy with his system. In the earlier session it was suggested that such cases should be presented during the chat session with reasons. After approval of the members the changes can be made. As only three of us were present the decision to change was postponed. Murthy suggested that in the meanwhile, the changed codes be posted along with the reasons for change so that the members can assess the case and in the next meeting take a decision. All the creators of the systems can therefore present their cases if change is needed. The modified code along with the reasons should be posted latest by Thursday so that on Saturday it can be discussed. As Sanjay has suggested continuation of the test for two weeks, the review of changes can be taken up next week.
THEREFORE ALL THOSE WHO WANT TO MAKE MODIFICATIONS TO THEIR SYSTEMS MAY POST THEIR MODIFIED CODE WITH REASONS FOR CHANGES BEFORE THURSDAY.
The session ended with Murthy's concluding views and a request to rpc to post the model reporting format.
The next session will be held on Saturday, 23-12-06 at 9:30pm. Please be present at the appointed time. The discussions will be held in 'Sysimp' room.
Regards
-Anant
