btanalysis ' s Posts - lapu 2

KeltnerPRO - Jared in Tirdzniecības sistēmas May 09, 2015 at 02:27

If you further look at the historical NFP reported vs. revised figures, you'll see that either the US Bureau of Labor Statistics is using some horribly inaccurate data collection methodology or it's trading made-up changes to made-up numbers: http://blogs.reuters.com/global-markets-forum/2014/04/02/will-nfp-surprise-look-at-historical-data-revisions-vs-forecasts/As for KeltnerPro, if you exclude its January 15 USDCHF spike which doubled this account of the vendor yet blew his other one with FXChoice, you'll find this bot's trading unprofitable since last year's November.

KeltnerPRO - Jared in Tirdzniecības sistēmas May 08, 2015 at 15:33

I was expecting a cascade of trades after today's NFP but the bot entered none at all on mixed data: despite the nearly consesnsus reading for this month, there has been a downward revision to mere 84K for March. With no strong recovery in sight, I expect the losses to continue for the remainder of month and into June. I will leave my bot on for the time being unless and until they will grow critical.

KeltnerPRO - Jared in Tirdzniecības sistēmas May 08, 2015 at 10:10

I had losses, just like the vendor. Anyway, they're largely irrelevant in the long run - today's NFP release is what will determine this bot's prospects for this year. If the NFP stats are weak, there won't be many good breakouts, whereas both the US and the EU economies will find themselves in a simultaneous slowdown for the months to come.

KeltnerPRO - Jared in Tirdzniecības sistēmas May 08, 2015 at 07:37

Losses for KeltnerPro have extended further this week. Today's US NFP will be its make it or break it moment - if it meets or exceeds the consensus expectations, there's hope for this bot; if the figure underscores the expectations yet again, it will mean there's no US economic recovery - then say good-bye to the Fed rate hike this year and a divergent central bank policies between the Fed and ECB, so much needed for a successful performance of this bot.

KeltnerPRO - Jared in Tirdzniecības sistēmas May 06, 2015 at 12:11

Which one?

KeltnerPRO - Jared in Tirdzniecības sistēmas May 05, 2015 at 22:26

Unless your BT was done on tick data, the earlier years might be largely inaccurate due to compressed MetaQuotes data from back then. Also, there wasn't a divergent central banks' policy between ECB/BoE on one hand and the US Fed on the other one, like there's now. Breakouts are best not only when Europe screw it up big time yet again but when the US situation improves at the same time (and vice versa). Run it since fall 2008 (when the global fun started) and you'll see fall 2012-fall 2013 as the longest stagnation period due to the Greek drama Act II on one hand and the US deb...

KeltnerPRO - Jared in Tirdzniecības sistēmas May 05, 2015 at 00:55

The current EU losing streak started on March 25. From there on, it has been only losing days on EU. On multi-year EU backtests, 2012-2013 has been the worst period for this bot. On a backtest over that specific period, the longest single EU losing streaks lasted since September 17, 2012 till January 02, 2013 (that's three and a half months) and since April 23 till June 18 (that's almost two months). If memory serves me, it was also due to the Greek drama back then. Difference between then and now? Back then, Greece wasn't facing a tough choice to either pay back the IMF (not doing...

KeltnerPRO - Jared in Tirdzniecības sistēmas May 01, 2015 at 00:15

Lastly, here's a SNB-adjusted (excluding the January 15 USDCHF event) ForexGermany's chart with the connected minimums/maximums which shows the recovery even more overdue for a whole month:http://www.myfxbook.com/files/btanalysis/ChartExtremesForexGermany_6l.jpg

KeltnerPRO - Jared in Tirdzniecības sistēmas Apr 30, 2015 at 23:46

For the above reasons, it would be more logical to use a live forward test's chart where the MM was not manually interfered with. One example of such a forward test of this bot is that by ForexGermany at http://www.myfxbook.com/members/forexgermany/keltnerpro/983876 where the EA has been run continuously on the recommended five currency pairs with the default settings from the very start. That live test's graph also suggests the bot's recovery is just about overdue. Yet how reliable is the very method of connecting a forward test's chart's three upper and lower extremes to ...

KeltnerPRO - Jared in Tirdzniecības sistēmas Apr 30, 2015 at 23:30

The reason I'm asking about the next maximum/minimum prediction methodology is as follows: if we remove the January 15 USDCHF once-in-decades "big swan" jump (which doubled this vendor's live account yet blew the other one), the chart will show that the bot's recovery is just about overdue. But then, again, the vendor manually more than doubled a trade's lot value from an average of 8.5 lots under the automatic money management before March 4 to a fixed value of 19.30 lots after March 4. So that makes me wonder how relevant the charting of lines connecting minimums/ma...

KeltnerPRO - Jared in Tirdzniecības sistēmas Apr 30, 2015 at 22:48

Hi BigSteve. What is it about connecting the three lower and upper extreme points of a multi-currency forward test chart to determine the next minimum/maximum? Could you, please, describe this prediction method in greater detail?

KeltnerPRO - Jared in Tirdzniecības sistēmas Apr 30, 2015 at 14:12

What's daunting here is an utter lack of any response or reaction from "Jared" with his opinion on the prospects of his own bot.Remember, your objective with this EA was to make some money trading while his objective was to sell as many copies of the EA as possible. These are two very different objectives. Perhaps (and I'm assuming here), his bets are on a major breakout or a series of breakouts in the near future to make up for this month's huge losses due to his switch to an excessive manually entered fixed-lot trading. What can get things back on track with the Keltner...

KeltnerPRO - Jared in Tirdzniecības sistēmas Apr 29, 2015 at 19:31

-30% after April already. Got to cut down on the risks or not too long before you're bust. Personally, I'll drop this bot if May turns out to be a second consecutive losing month after April.

KeltnerFlow in Tirdzniecības sistēmas Apr 25, 2015 at 22:50

I can't copy the EA file from the ZIP archive without entering a password. What's the password?

KeltnerFlow in Tirdzniecības sistēmas Apr 24, 2015 at 12:49

Today, the original KeltnerPro EA took losses because of a late trade opening due to slippage. KeltnerFlow has demo-profited. Such a feast!

KeltnerFlow in Tirdzniecības sistēmas Apr 22, 2015 at 13:34

Better yet - pick the biggest single breakout over the backtested period. Set the bar limit to the size of that breakout. PF on backtest? Intense! Accuracy? Extreme! Chances of a similar outcome on a forward test? Next to none.

KeltnerFlow in Tirdzniecības sistēmas Apr 22, 2015 at 08:40

I have my doubts about the relevance of aiming to increase the PF at the expense of the number of trades when it comes to this type of strategy. Markets don't treat the same things the same way over time. Just some years ago we would have wild rides on such Greece news while now the market hardly moves. So it's not unusual now when an incidental breakout does end up with trades hitting TPs due to something like a thin market while a perfectly healthy, logical breakout from the standpoint of the past soon reverses because the market players now realize in increasing numbers that both th...

KeltnerFlow in Tirdzniecības sistēmas Apr 21, 2015 at 19:53

That would hold true if the higher profit factor led to higher fixed-lot profits, too. That's something I fail to see from your backtests - rather the opposite takes place, like those meager 30% over 3 years on EU.

KeltnerFlow in Tirdzniecības sistēmas Apr 21, 2015 at 11:53

Please, let us know once that is done. I feel like I'm increasingly missing the point here - after seeing those 3 new 3-year backtests with profits much smaller than those of the original KeltnerPro bot. Especially, +30.25% on EU over three years. 😲

KeltnerFlow in Tirdzniecības sistēmas Apr 19, 2015 at 00:57

In particular, I would love to see your bot's 15-year USDCAD backtest - that of the original KeltnerPro bot is attached below and sucks beyond belief while you've nailed some stunning profits on this pair on your demo test right from the start.http://www.myfxbook.com/files/btanalysis/StrategyTester1.gif

EURUSD	1.07757	GBPUSD	1.25673	AUDUSD	0.66281
USDJPY	153.852	USDCAD	1.3659	USDCHF	0.9056

Pievienoja lietotājs forex_trader_67253