Spotting entry and exit points with heikin_ashi

anon46773462 · December 2, 2017, 8:19am

[quote=“captgrumpy, post:58, topic:122057”]
It seems we are not on the same page about some thing(s).[/quote]
I confess I was a bit taken aback by this statement at this stage! I really thought I had understood the mechanics of this method!

I didn’t see anything in your post that is different to what I thought I had understood, so I am a bit confused now!

Could you please tell me what exactly you feel we are on different pages about?

If you are referring to the way I describe my own trades so far, or my comments on the position entry timings then I plead “guilty as charged”. But these are just personal comments and not misunderstandings of the original method. I had thought that the idea here was for us to test and try this out and comment accordingly, not to just mechanically follow the rules.

Clint is doing the formal precision rule-based comprehensive test across a broad range of Instruments. I think it is good that others of us are questioning and exploring these rules - for example putting on new trades on a Friday close.

But if I have really misunderstood something that my other brain cell has failed to register then please put me straight with a specific explanation because i didn’t find it!

Thanks!

anon46773462 · December 2, 2017, 10:22am

@captgrumpy
I am compiling these rules into a doc. Can you confirm the following:

The spike in the WMA5 is only created and appears at the close of the following day’s candle, at which point the trade is entered. In other words, at the time when the trade is actually entered/closed the actual PTP is in the preceding day’s HA block. As in these examples. Here is the 28.11 USDJPY chart at the close of that day (no spike no changes):

but then on the 29.11 the block changed colour and the WMA5 changed direction thus creating the spike now showing on the 28.11 block, but the trade is actually entered at the close on 29.11 (the red circle) - and then we follow the WMA5 until it crosses the WMA12, which here occured in the orange circle:

Pont · December 2, 2017, 11:08am

When you open the position where will you get out of the market? What is your potential risk and profit?

anon46773462 · December 2, 2017, 11:18am

As I understand it, there are three alternatives:

if the 5WMA crosses the 12WMA then you stay in the position until the next PTP appears, at which time you can close or even SAR it.
if the 5WMA fails to cross the 12WMA then close it as soon as seems sensible (I don’t think we have talked too much yet about how long to wait for a cross, but I would think not more than the next 1-3 candles).
the usual stoploss at whatever level one has selected to place the stop. This is also a discretionary issue at present, I believe.

anon46773462 · December 2, 2017, 1:59pm

This NZD/USD is also a SAR, is it not?

Also, I think there is a very serious question here which is illustrated here on the AUDUSD!!!

I have not seen any instruction here about what “close” is being used for the LWMAs on the HA chart. There are two options: either use the actual closes from the normal candle or the HA closes! I have been using the HA closes as that seems to fit with the concept of using the HA smoothing candles in the first place. But we get different answers:

This is the 5LWMA based on market closes, which shows a PTP:

And this is the same 5LWMA based on HA closes - which does not yet indicate a PTP:

So @captgrumpy which close basis have you been using? Personally, I think this is an extremely critical issue here that affects both entry and exit decisions?

PS: Just as an observation, this particular pair looks to have been a nightmare over the past week with maybe three PTPs and with WMA crossovers (but the actual pips loss would not have been very bad)…

Falstaff · December 2, 2017, 2:15pm

Am I understanding that correctly Clint ? - Still positive overall including those “Spikeouts” ?

Falstaff · December 2, 2017, 2:35pm

That looks as though the price action would show either a retracement, or a potential retest of the previous swing low.

This is somethong, which we could only guess at and would need to study the longer timeframes and Support /Resistance evidence to assess the probability.

However, we are using Heikin Ashi to smooth the action and prevent us being whipsawed too often, plus the indication is that a double bottom has not yet been formed and this was a very volatile last bar. If it refers to action on Friday, we also know that there was a considerable news “Spike” which shows on many USD instruments, so my inclination would be to give it another bar to confirm what the market wants rather than to accept the price action based SAR at face value on that one.

Clint · December 2, 2017, 2:45pm

Correct. Realized P/L and unrealized P/L are both positive (showing profit).

Ten positions got stopped out on Friday. All the stops were trailing stops, so some of the stopped-out trades were profitable when closed. That pulled Thursday’s (negative) realized P/L into the black.

Unrealized P/L settled out at 175.6 pips, when the Oanda platform finally shut down after the Friday close – slightly more than I reported in my Friday metrics (previous post).

EDIT:

I have changed all the stops from trailing to fixed, so beginning next week, stop-outs (when they occur) will invariably result in losses. I hope that there will be fewer stop-outs, going forward.

It might become necessary to increase the fixed stops from 100 pips (away from the entry price) to 120 pips, or 150 pips.

For the next week, or so, I will run with 100-pip fixed stops on the USD/majors and the major crosses, and much wider stops (or no stops at all) on the two USD/minor pairs (USD/MXN and USD/CNH).

Stop-loss placement is more art than science, as you all know.

Falstaff · December 2, 2017, 6:16pm

Here’s the chart of the 1 hour Actual price - showing proxy volume as abnormally high for the first 4 hours of the move. There is some follow through towards teh end of teh day, but it was a Friday and I suspect some of the longs from the white bar attempted recovery were just closing out for the weekend.

Will it retrace ? I don’t know but I suspect it will. Who knows ? but I’d not be in any great hurry to go short on monday

Falstaff · December 2, 2017, 6:37pm

Stop loss placement errors have cost me more real money then any other cause - except possibly spreads over the years.

anon46773462 · December 2, 2017, 6:43pm

Hi!
This is yet another example of the question about what close should we use for the LWMA 5 and 12 - normal candle closes or HA candle closes (the HA block is, of course, the same in both cases).

If we use the normal candle closes then the method already gave a sell signal on the close on Friday:

But if we use the HA candle closes (which I prefer) then the LWMA 5 was flat and therefore not a clear sell signal yet:

This question has existed since the early posts of this thread and this is the third example of how this can give totally different signals on occasions. I am still not sure what is @captgrumpy’s thoughts on this…

@Clint said in post 11 " In the two charts posted below, the WMA’s on the H/A chart use the H/A closing prices, and the WMA’s on the normal chart use actual closing prices" but the example in post 65 regarding Friday’s AUDUSD seems to suggest that the actual closes were being used.

anon46773462 · December 2, 2017, 7:01pm

[quote=“Simple_Simon, post:41, topic:122057”]
I was wondering, is this swinging from big to small profits due to the method itself or because of possible excessive correlation or imbalance in exposure in some of the pairs we are looking at here? [/quote]
This is another issue that I think would be worth addressing at some point. Whilst it is good to test the method with a broad range of Instruments, I think it would be prudent to consider some rules or guidelines when designing an actual portfolio for actual trading.

For example, there is a danger of excessive exposure risk if the range of Instruments is too heavily biased towards, say, USD or GBP or EUR or JPY, etc.If there is a major move in one of these it will have a distorted impact on the overall profitability - good or bad!

Also, whether it is worth including commodities and stock indices.

Another currency basket consideration would be whether it would be worth selecting pairs based on some kind of strong/weak analysis such as Dennis3450’s thread?

But this is not an urgent matter at this stage, but maybe worth thinking about a little if planning to actually trade this method…

captgrumpy · December 2, 2017, 7:24pm

Hi clint–About WMA base–no choice on this platform,but I would think it is market based since your chart with market base showed a PTYP while the other didn’t

about the AUD-USD— my chart shows a PTP end of 30/11 so it looks like trades could be made at closing on block 1/12,but that is
Friday so trades held over 'till Sun closing That would be close the sell–open a buy—this chart at this moment is not good, too choppy an unsettled --I would be inclined to hold all trades until chart shows better movement.
the NZD-USD looks almost identical–PTP close of 30/11 trades
made closing 1/12 held over until Sun or Monday-- again a poor chart to work with
I think a lot of problems stem from the fact the charts I am using
(oanda Fx Trade program)set at maximum resolution are different than the charts respondents are posting.With my very limited knowledge and experience any other charting system totally confuses me–If I knew how I would submit a chart or two so everybody could see how I come up with PTP’s, entry/exit points,etc.
I am extremely pleased there is so much interest–it seems most think the plan has some merit,but there are a lot of bugs to be ironed
out–I am being overwhelmed by all the inquiries–I,d like to respond to all,but may not get back to everybody–

On submitting records of my trading -it is pretty sparce -only have one active trade-USD-JPY opened at B 11185–on 29/11-- as of 1/12 is 11260 +75 pips --Also had a trade
on EUR-USD -29/11 S at 11859 --had to close at 11919 loss of 60 pips (Entered as a sell in error,supposed to be a buy).Will be doing more trading–will post results

Clint · December 2, 2017, 7:37pm

I think you meant that reply for Simple Simon.

Falstaff · December 2, 2017, 7:56pm

I don’t seem to be able to get the HA based MAs on my charts ! So my charts show the suignal.

But taking a closer look, we are now at the bottom of a range which has been trading for the last 5-6 weeks. It powered down there on Friday from the top of the range and sits atop major long standing support / resistance.

CAD has performed much the same against all the other pairings I checked, on Friday, some of the movement on other pairs started on thursday.

captgrumpy · December 2, 2017, 7:58pm

Hi Pont—I trust you have seen my post–Entries are made at closing of the next block after the PTP-( Potential trade point–on the original post this was called a TP–trade point) was spotted(At max. resolution,the candles become solid blocks!) Exit is normally at the next PTPBut if the WMA 5 does not cross the WMA12 the trade is closed and then wait for the next PTP to open a new trade. As for expected losses I don’t expect any amount–I don’t set up a formal stop loss–if the trade is closed because theWMA5 did not cross the WMA12 there could be a small loss,but here there is usually a bit of pip gain before the closure–once the WMA5 crosses the WMA12 the WMA 5 is hollowed until the next PTP when the open trade is closed.,so it is impossible to predetermine the profit as that depends on how long the run is–hope this helps–captgrumpy

captgrumpy · December 2, 2017, 8:13pm

Hello Bobcat----Glad to see you are experimenting with the plan–this is what we need to finally come up with a reliable trading plan
Using both the WMA2 and WMA5 definately gives one more confidence making,and exiting trades Good luck with your new
trading strategy–captgrumpy

Falstaff · December 2, 2017, 8:34pm

Yes, the neccessary parameters for this method go against several mantras of “the narrative” and is similar in this respect to “Trend following” (like “tight stops” and unknown “Risk reward ratios” )

I shall certainly use stop losses if I do trade this, but will probably assess each one according to the past history of the pair under consideration. To that extent, I feel it will require an amount of work each time a trade is entered. For example if we were to go short the USDCAD mentioned above, the logical stop loss would be away above the Thursday highs, but that would be huge and if it were to go back that way, it would probably take several days and would be worth a trade in its own right. Almost undoubtedly we would have a PTP by then anyhow.

Xanaphlaxes · December 3, 2017, 4:38am

Last time I checked, which was admittedly years ago, OANDA’s demo feed didn’t match the live feed.
The demo accounts are great for learning your way around the platform. Just FYI that demo data might be misleading if your goal is to figure out whether or not a strategy will be profitable. Demo data policy may have changed, but I would check to be sure. Some brokers do not guarantee demo data to match live.

OANDA allows extremely small lot sizes for real trading though, so it is easy and low risk to test with live account and small trades and then scale up.

Clint · December 3, 2017, 5:50pm

There has been a lot of discussion in this thread about the discrepancies between price feeds based on standard candlestick charts versus price feeds based on H/A charts, and the resulting discrepancies between their weighted moving averages.

It should be clear to everyone following this thread that different charting packages will produce different signals if (1) they use different times to open/close daily candles, and/or (2) they use different methods for calculating the weighted moving averages (WMA’s) displayed on their H/A charts.

In my test of this strategy, I have tried to replicate the methodology outlined in post #1 (with the exception of the number of pairs traded). Captgrumpy said that his “plan” – what I call his “strategy” – uses Oanda’s H/A charts. So, I set up my test of his strategy using Oanda charts, so as to see what he is seeing.

If all of us were using Oanda’s H/A charts, all of the confusion about which WMA’s to watch would go away.

I think we should examine the Oanda charts in order to understand exactly what they display. Then, those persons who choose to use other charting packages will have the Oanda baseline to compare their charts to.

• Oanda charts open and close daily candles at 1700 in the GMT-5 time zone – that is, 5 pm NY time. This is true for every Oanda customer, everywhere in the world (although captgrumpy does not believe this). For traders in London, where the time zone is (currently) GMT, the Oanda end-of-day is 10 pm, which is precisely 5 pm NY time. For traders in Sydney, where the time zone is (currently) GMT+11, the Oanda end-of-day is 9 am, which is precisely 5 pm NY time. And so forth.

If you use a charting package in which the end-of-day is something other than 5 pm NY time, and if you cannot adjust your charts to conform to 5 pm NY time, then your charts cannot and will not look like captgrumpy’s, or mine.

• Oanda uses actual closing prices to compute indicators, including WMA’s, on all of their charts, including H/A charts. Therefore, the WMA’s plotted on standard Oanda candlestick charts are identical to the WMA’s plotted on Oanda H/A charts, even though standard candles differ considerably from H/A candles.

• The open, high, low, and close in Oanda H/A candles are calculated using the commonly accepted formulas. As with all other H/A charts, the candle bodies (representlng the open and close, and the range between them) differ considerably from the bodies of standard candles. There is one exception to this: the most recent price in the most recent Oanda H/A candle is always identical to the actual price. This remains true when the most recent price is the last price for the week. That is, when the Friday H/A daily candle closes, the close depicted in the body of that candle matches the true closing price for the week. You can see this in the charts displayed below.

The following images will illustrate these features. (Click on the images to enlarge them.)

Here is a standard Oanda candlestick chart of EUR/USD, showing the most recent 50 candles and the WMA(5) and WMA(12).

Here is an Oanda Heikin Ashi chart of EUR/USD, showing the most recent 50 candles and the WMA(5) and WMA(12).

.

Here is a side-by-side comparison of the standard candlestick chart and the H/A chart, showing the most recent 15 candles and the WMA’s.