TSF Mavin

I've done a quick look back through August of last year with regard to US Unemployment announcement Fridays. Always the first Friday of the month. In some months the GBP rate announcement happens on Thursdays of that same week. Other times it is on the following Thursday. This is because the GBP Rate announcement is made on Thursday of the first FULL week of each month.

When both announcements fall in the same week, the odds of Mavin making a bad call are increased on the Friday unemployment news day. But in months where the announcements fall in different weeks, the unemployment news day is not especially bad to trade... just be sure to close your trade no later than an hour before the news or you are at the market's mercy. If a trend develops after the news it is probably okay to trade, but there seems to be a little higher probability of a loser then... And if there is no appearent trend, just back away from your computer for the rest of the day.

This week we see both announcements in the first week so the word to the wise is don't trade tomorrow... ... hmmm... Will I take my own advice? hmmmm....

So tomorrow, which is often a very flat day (before US numbers post) could be a tough trading day too. But I think there is so much news out today we will probably see a trend into at least the early part of tomorrow, so I'll be looking for my entries much later today into the early AM tomorrow -- At least with the Mavin. I have no idea what the London open will bring, so I'll probably stay away from anything there on Thursday. But I do expect the US market open is going to leave us largely directionless, so I think that might be it for me for the week, unless I don't heed my own advice... At least with the Mavin.

Ah shucks.... After all my warnings and "expert" analysis, the market throws and easy first win, and if anyone dared, is on the cusp of closing a second win... will it, or won't it....

There, see how smart I am. The market was so well behaved a 10 year old could have traded this system for.... yes, two wins. I told you I'm not always that smart. But you know, I wasn't awake for most of those festivities. Look at that news. Everything either met the expectation or disappointed. If any one of those, or say two of those had turned out to be positive surprises, anybody who traded the Mavin would have been Barney the Dinasaur fighting the Great Ali. I might not be looking so smart today, but change those elements and I'm Einstein.

If I could get by with no sleep, here is what I hope I would have done... First trade - No way. The first signal came at 0130 GMT and would have filled on the next bar. The score didn't come until 0830. Those 6+ hour holds will make you old before your time. Too much risk there with all the news and the Rate Statement to boot. But after all those numbers folded out, you get the retrace to reset. Since all the news was in agreement for market direction, let me tell you, there was no fundamental reason for the market to retrace. But it did. That's profit taking to a point, and stop hunting for sure. Once the board is cleared of the easy money, back it starts to come.

The Mavin will give good signals as long as the market remains within reasonable limits. From TSF Mavin's perspective, this pull back wasn't enough to change it's vote on the actual sentiment of the market, and it signaled a trade as soon as the market confirmed it was returning to the trend. With all the fundamentals to that point in agreement, and with the last bit of it being US news which left no reason for the reverse, the new signal bar is okay. Had I been awake I like to think I would have taken it.

If another signal comes along anytime before Friday's London open, then I'm going to be in. If I'm doing this completely mechanically though (say I had an EA), I would have turned it off and missed these trades. That is what you lose when you have an EA. In any given time span I'm going to miss a fair number of trades. I just can't put in 24 hour shifts to watch, and I've got another job so I can only move my sleeping time so much. Most of you know what I mean. But here in Hawaii the alignment in time means I miss most of the London session. So believe me, when you're trading THE BEAST... you try to choose when to place your risks as intelligently as possible. Trading in and around the London open last night would not be smart with this method. Looks good on paper today... but in the long run, you'll be torn and tattered.

Stevewide has the numbers though. In theory his calculations left out these announcement days. If that's true you see the numbers, skip these days and you are a long term winner (or so it seems at the moment). If he mistakenly counted trades on these days, then I'll reconsider the advice. It would be interesting to see what the numbers would be if we did ignor those days. I'll happily go back and check, but I don't know for sure if he tossed the days out or kept them in. If leaving them in comes out better than tossing them, well TSF Mavin is a lot smarter than I give it credit for... Shoot, I'd say bring on the EA, baby. But if it does this well without those days, or maybe does well even with those days -- wow

Thanks again Stevewide. Okay, so here I sitting & waiting for the next opportunity. Hope one comes before London... and I'll watch carefully if I'm still in when London open comes... I expect not much movement, but who knows, market doesn't live to meet my expectations, if you know what I mean. I'll close my position early if I'm in doubt at that point.

One last thing, in general the system ignors the news. The only thing we are iffy about is currency rate related news, and even then almost exclusively the GBP. I'm not so much concerned on any given day what the news is, but I do note that it my speed or delay my take profit, or hasten my stoploss. We are going to lose sometimes. Never like it when it happens, and when there appears to be a good reason for us to have stayed away that is really just emotion. Because with an 85% win rate, I promise you, if you look back, on 70 percent of the winning days you'll see red news items that would make you think you should have held out. In general, with a 70 pip stoploss, those news items mostly just don't matter. The trend is the trend. Mavin finds trends and is reluctant to change. When enough things align to give a signal, even on very unvolitile time spans, and with the ribbon lines whipsawing all over the place, there is enough rythm in the market to signal a profitable trade a high majority of the time. Despite whatever the news may be. But the rate statements can move a market for days, and initial reactions can be very unrythmic (kind of like my dancing). Even seemingly unrelated pair's rate statements can color the mood in the beast for a few hours. You'll lose some of those, and win some of those. Can't jump out for every rate statement related news bit though or you'll barely ever be in. Can't win if you're not in. But you'll always lose some. Sit it out, and you'll never lose -- and some days you'll look like a genius -- and other days you'll look like ----- me, calling last night ....

It's okay. 38 Pips per day average can make you rich pretty quick, even with a very small account. If I knew today would be a freaking 204 Pip move, I could get rich a lot quicker. --- But I knew I didn't know. I never know. In fact what I really know is whenever I think I see a 200 pip move coming, nobody is worse than me at picking it. I've got the zeroed accounts to prove it!! So I try to settle for reasonable pieces off of the moves. Wish I'd been awake at 3:30 AM here in Hawaii though. I'd be 30 pips richer, almost certain of it.

Best of luck to all, and be careful on Friday. US Unemployment number in this economic situation is almost as important as the rate statements.

Qman:
Enjoying your observations and comments. I updated my post to include the fact that I counted trades regardless of what was happening that day.

Still learning, listening, observing but I like the quick 30 pips today I made.

I set a pc timer to ding every 30m so I remember to take a peep.

Thank you very much,

Steve

Hey, Stevewide... Did you count xmas & new years weeks too?

Maybe I should have you set your alarm to ring 5 minutes before each half hour. If there is something to consider, you'll then call me either at my office or on my cell phone....

.... Cheap man's EA...

Additional Analysis........
Consecutive losses: twice lost 2 trades in a row(November 5 and November 9-10, 2009 and once lost 3 trades in a row(November 13-15, 2009)

Found an error in analysis resulting in 1 additional win and 1 less loss...

Hey Man:

Get your own free timer here..........

http://www.nestersoft.com/timeleft/download.shtml

Thanks for the link, but I don't see where you can get them to check for signal bars

Seriously though, did you count xmas and new years week trades during your analysis. It would seem not as I believe there may have been consecutive losers during that time. But I'm just checking. With your answer I'll be looking to redo my posted numbers from yesterday. It's going to be smaller than the 38+ pips, but who's worried about that?

Thanks again for your continued research. By the way, November 5th was the GBP rate statement day for that month.

Christmas week includes 6 wins , 3 losses...
Week after Christmas was 4 wins

This data is included in my analysis.

Steve

Thanks Stevewide. I think I have it now. So, updating through the close of Thursday, the record is 119 wins, 21 losses for 2100 net pips. No trading days or trades skipped. Winning percentage now approaches 87%.

2.044 Trades per day
1.78 Wins per day
0.313 Losses per day
31.34 Net Pips Per Day

Worst run, perhaps 5 losses in 7 trades.

I will update the running statistics and post over the weekend. Stevewide your contribution has been very helpful. Thanks.

2 Losses and 1 Win yesterday. The numbers below assume 70 pips lost on each losing trade which sometimes isn't the case.

1.57 Trades per day
1.31 Wins per day
0.25 Losses per day
28.84 Net Pips Per Day

I thought it would be a tough day. I also thought there might be something that shakes free late in the day. That's pretty much what happened... But it was too late for me to jump in cause of work. Hope somebody cashed in.

I have analyzed October. For anybody thinking this method is looking "grailish", I urge you to take a good hard close look at October 2009. The good news: For the month, the average losing trade only lost an average of 58.82 pips, not 70 pips. But that is about the extent of the good news.

Total Trades: 58
Winning Trades: 35
Losing Trades: 23
Winning Percentage: 60.34%
Winning Pips: 1050 Average: 30/Winning Trade
Losing Pips: 1353 Average: 58.82/Losing Trade
Net Pips: -303
Biggest Closed Trade Pip Drawdown: -483
Worst Losing Streak: 4 in a row
Worst Losing Run: 12 Losses 7 Wins

A pretty dismal month by any measure. The question on the table is, would there have been anything to indicate this was such a poor period to trade. Actually, the first trade of the month was a loser, but was followed by 8 wins. Then came the 12-L/7-W run. The trouble began on 5 October. The Mavin didn't return to Mavin-like accuracy until the last 16 trades of the month, or roughly around the 21st of the month. Let's all focus on that period and see if we can understand why it was such an incredibly bad period for the method.

Overall results for the method are a little fudged right now because I don't have any accounting for partial losses since 1 November. We are presuming all the losers since November have run to a full 70 pip loss, and I'm pretty sure that isn't the case. Regardless, this is what I have for a tally:

Winners: 155 Losers: 46 (The number of losers just doubled by adding Oct)
Percent Winners: 67%
Net Pips: 3821
Trades/Day: 2.2
Winning Trades/Day: 1.7
Losing Trades/Day: 0.5

Anybody have any thoughts?

I may have failed to apply one of the qualifying filters in my analysis of October. Specifically, I'm not sure I refused trades within 20 pips of an extreme daily price. Rule says if your entry would be within 20 pips of the high or low it does not qualify. I noticed as I was calculating September. I hope to post numbers later today.

I have not yet double checked those awful October numbers.... But here are the awful September numbers.

Winning Trades: 30 Losing Trades: 24
Winning Pips: 876 Losing Pips: 1350
Net Pips: -474
Avg Win: 29.2 Avg Loss: 56.25
Avg Trades/Day: 2.45
Most Consecutive Losses: 6
Worst Losing Run: 8 Losses 2 Wins

Pretty Dismal. Again, a good start to the month (through the 3rd), 10 Wins and 2 Losses. Then the wheels came off. Another good Mavin-like run from the 11th to the 16th, 6 Wins and 1 Loss. Aside from that an umitigated disaster.

So it would seem the performance since November is possibly not characteristic of the overall odds. I do believe Sep and Oct 2009 were the worst performing months in the past year, but again, I haven't looked much beyond the first trade of the day -- but using that as an indication, I knew Sep and October were tough months. I'll try to get to checking October over again today, but no promises. Getting a bit discouraged with what I'm seeing here. Unless I can figure out some fundamental reason that flags the bad weeks, or find some additional filter or indicator, this system would seem to be too problematic. If I can't think of something in the near term I'll probably be drawing this thread to a close.

I haven't been keeping track, but I have noticed several times when the Mavin has indicated/signaled a trade shortly (5 hours or less) before the close of the week on the IBFX feed (which shuts down a couple hours earlier than many brokers) and when those filled before the weekend there was a nice profit to be had on the open the following Sunday. But if true, that would be a different method altogether, and would require a new thread if I were to discuss it beyond what I have here.

Pips to all.