Extreme Deviation Automatic Price Targeting - "X_DEV" - Bitcoin (BTC-USD)

[AIMAGIC]

Hi Myst,

I am concentrating on daily spreadsheets currently.

Do you know how the XDEV program generates its “buy forecast” and “sell forecast”?

To recap:

The XDEV sheet uses 3 sma price information and the XDEV multiple to generate the HIGH % and LOW % DEV bands to “forecast” the price levels at which to trade the next day. This is historical data.

The question we are asking is how best to generate the amount to BUY or SELL?

Currently, the amount to BUY or SELL is calculated with reference to prices inputted the same / or next day.

While the levels at which to BUY or SELL are “forecast” or forward-looking, the amount to BUY or SELL are calculated with reference to information in the spreadsheet that is current and is generated in “real time”.

As you understood, this makes it impossible to input BUY or SELL orders in advance.

You say you are nearly always at home when you trade. Can you say what that looks like?

Does that mean that you are watching the price and if the price reaches either the HIGH % or LOW % band targets you enter the (current) OHLC information into the sheet to generate a BUY or SELL amount?

Does that mean you will in instances have to enter current / ongoing OHLC information before the end of the trading day?

This is fine, as you say, but means you have to be watching the charts to enter the current OHLC price information into the sheet, which cannot be done in advance.

However, doing this means the amount you buy or sell will be calculated using the most up to date price information.

It seems that the “cost” if we want to set limit orders the day before is that we have to also quantify the amount to BUY or SELL using data from the day before.

Let me know what you think after you have have given thought to incorporating that feature back into the spreadsheet.

You say: “Use the high or low from the period before to let it forecast the buy or sell amounts.”

In terms of the spreadsheet this means changing the formula in the BUY Market Order column from:

=IF(D10<I9,ABS(AA9*P10)-$M$4*(AA9*P10),”0”)

to:

=IF(D10<I9,ABS(AA9*P9)-$M$4*(AA9*P9),"0")

and changing the formula in the SELL Market Order column from:

=IF(C10>H9,(ABS(AA9*P10)*-1)-(AA9*P10)*$L$4,"0")

to:

=IF(C10>H9,(ABS(AA9*P9)*-1)-(AA9*P9)*$L$4,"0")

Essentially, this introduces the one day lag.

I have shared a sheet with the change above. The results are still good. No doubt you will have a better grasp of the drawbacks of doing this.

I can send over another sheet with the same changes if you want to work on a specific play.

Best regards,

Alex

[Myst]

Ok...now I see the issue. I hadn't considered it before because the program, as you can see from the screenshot below, has the "buy forecast" and "sell forecast" feature which tells you the amounts to buy or sell in advance.



I'll need to give some thought to incorporating that feature back into the spreadsheet. Since I am nearly always home when I trade I have the benefit of not needing to place trades in advance. I can input the figures into the sheet and let it tell me the amounts in real time. I only use daily and 4 hour spreadsheets so I don't have the same issue as one that uses weekly or monthly data.

Until I have figured it out you can just input current data into the sheet and use the high or low from the period before to let it forecast the buy or sell amounts. It may not be exact but it will be close enough. I wouldn't use the previous periods XDEV factor. It would be more accurate to use current data especially if you are working with weekly or monthly sheets.

[AIMAGIC]

Hi Myst,

In the example shown in the screen shot, the question, then, I suppose, is should we not use the DEV factor of 10.2% from the 7 12 21 rather than the DEV factor of 8% shown 8 12 21?

Unless I am mistaken, currently, the DEV factor is calculated using price information relating to data on the same row. This information would not have been know the day before.

The amount bought or sold would need to be calculated with reference to the DEV factor from day before.

So, if I were to place limit orders in advance at the price levels indicated by the HIGH % and LOW % Dev Bands on the 7 12 21 - i.e. $175.74 and $139.66 - I would have to use the DEV factor from the same row, in this case, 8%.

Best regards,

Alex

[Myst]

Yes…. On the last column ( column V) on Aimagic MSTR sheet I have the % buy or sell portfolio value calculation

[Myst]

Look 2 columns over in the “sell market order” column. If you are using the same amounts in the spreadsheet then you would have sold $6466.66.

However I also have a column (I think column AE) on the XDEV sheets that calculate the % buy or sell of your portfolio value. So if the dollar amount you have differs from the sheet you will still know how much to buy or sell.

This could be incorporated in the Aimagic sheets as well. Not at my computer now so I’m not sure if I have done this already.

[AIMAGIC]

Hi Myst,

Oh wow. I have definitely been reading the spreadsheet wrong. That makes all the difference! LOL

I see now how 7 12 21, the High % DEV BAND indicated $175.74, which would allow me to enter this price as a SELL LIMIT ORDER in advance.

I could do the same for a BUY Limit Order using the Low % DEV BAND price indicated as well.

Now, forgive my asking a really stupid question, but where, then, is the size of each LIMIT ORDER also indicated in advance please?

Alex

[Myst]

Sorry for the delay in response....I was traveling the last few days.

The sell target of $175.74 was reached on 8/12 which means if your limit order was in place as indicated on 7/12 you would have sold.

I think you are misreading the spreadsheet signals. The sheet gives you the targets in advance....not for "the day after".

Does this clarify the situation?

[AIMAGIC]

Hi Myst,



This screenshot comes from an NRGU play.

8 / 12 / 21, the close was 170.81 and updating the spreadsheet indicated selling almost all STOCK at a price of $175.74 the next day.

However, the next day, that target was never reached and has still not been reached so you never got the opportunity to sell at the price indicated on the sheet.

There have been 2 BUY signals since then, as you can see. If you had set a limit order to SELL or were waiting for the price indicated on the sheet to SELL, you would be unable to follow the trades on the sheet. Effectively, 100% exposed, waiting for the opportunity to sell.

Have you come across this kind of situation?

Kind regards,

Alex

[Myst]

“The greatest shortcoming of the human race is our inability to understand the exponential function.”
-Albert A. Bartlett-

A good trader can achieve good returns following a good system.

An exceptional trader can achieve exponential returns following a good system that has been optimized.

[Myst]

Looks great Alex! Nice job....thanks much!

CELO is really crushing it vs. Buy & Hold!

In regards to your other question I missed, for the spreadsheets any amount of starting cash can be used to optimize the play. I tend to have at least $5k in cash allocated in real trading scenarios for a minimum, but you don't need that amount.....it's just a personal preference.

Also, not sure if I mentioned this before, but I have spreadsheets for Lichello's 10-8-5-4-5-8-10 scenario and X_DEV's 10-8-5-4-5-8-10 scenario.

I will share them with you, so they will be there if you are interested.

[AIMAGIC]

Hi Myst

I've shared an amended version of the CELO sheet with you in which I added an "ABS(...)" (enclosed in a set of brackets) in the formula in the SELL Market Order column.

Now it looks like this:

=IF(C13>H12,(ABS(AA12*P13)*-1)-(AA12*P13)*$L$4,"0")

It looks like it has sorted out the problem.

Can you tell me if you are happy with that?

Best regards,

Alex

[Myst]

X_DEV CELO chart

May be a buying op here:

[Myst]

Upon further review CELO is an acceptable play for X_DEV. I think I had the symbol mixed up the first time I looked at it on Yahoo finance. So apologies for the misunderstanding.

I already multiply by -1 in the formula I have for the SELL market order which is why I am puzzled by the positive result in some cases. But if you can come up with a conditional formula that works then by all means, have at it!

Thanks again for your continued interest and efforts. Very much appreciated!

BTW, I have started 2 new sheets for a 4 hourly time frame study on Bitcoin and Ethereum. If you or anyone else is interested let me know.

[AIMAGIC]

Hi Myst,

With regards, the positive values in the SELL Market Order column, could we add some conditional formatting which multiples all cells with positive values by -1? That way we could eliminate all positive values from this column? Would that work?

Best regards,

Alex

[AIMAGIC]

Following on from my previous message, please ignore the moving averages I have on the BTC chart below, I overlaid CELO and BTC. Their "volatility" on a peak to trough basis is almost the same. CELO might even be a little more extreme.

Unless I am mistaken, CELO is almost as volatile as BTC and offered one more chance for compounding returns in the "up / down" cycle it showed in August 2020 which was not mirrored in BTC.

[AIMAGIC]

Hi Myst,

Thanks for re-optimising the sheet. In future, if I encounter a positive value in the SELL column, I will manually change those cells to negative values.

Apologies if I used the wrong term. Disregarding the absolute values of “2” and “4”, by “high volatility” I meant repeated “up and down” movements where the stock loses and then regains over 50%.

My impression is that it is these “volatile” up and down movements that allow XDEV (or AIM for that matter) to outperform B&H.

In the months shown on the sheet from 14 5 21 to 8 12 21, the asset is down -18.11% whereas XDEV has returned 187.76%.

On its own merits, that is an astonishing return in approximately 7 months and, as you point out, for a ranging asset that has gone nowhere, astonishing outperformance.

CELO stood out to me because the “up / down” cycles seemed both frequent in time (more opportunity) and “extreme” (greater than 50%), I thought, giving XDEV more opportunities to compound returns.

The chart reminded me of the Robert Lichello “10, 8, 5, 4, 5, 8, 10” sequence.

Since it isn’t certain there will always be long-term, upward-sloping trends in asset prices, the ability of XDEV to compound gains trading volatile assets by selling high and buying low is what appeals to me.

Have I misunderstood? I simply don’t have as much context or familiarity with what will make good or bad assets as you do.

If I start your optimisation with 200% CASH the return drops a small amount to 184.16%. If I drop the starting CASH to 100% (I.e. 50/50 Stock/CASH) the return drops to 21.07%.

In practice, would you start a play with $5,000 in CASH and $253 in STOCK?

Do you tend to start all your plays with around $5,000 in CASH? Or, do you simply try different CASH values in order to achieve the highest return?

If I increase the CASH to 3000%, the return increases only to 189.61%. Such an increase seems to have a diminishing effect on the return.

The orange cells in J5 and K5 (if I have calculated correctly) annualise the total return. They take the total profit or loss in the look-back period for XDEV and for B&H in cells J4 and K4 and turn these into equivalent compound annual growth rates (CGAR). I am not sure yet how useful this will be.

Alex

[Myst]

I've re-optimized your CELO sheet. I still haven't found an answer to the positive value in the sell market order column. I have manually changed those cells to a negative value and signified the change by putting a border around the cell.

I'm not sure CELO is a good candidate for X_DEV. It has only gone from about $2 to $4 in the last 5 years. Not my definition of high volatility. Also the daily space between high and low price is inconsistent. Sometimes the dev factor has stalled below 5% though other times it soared very high. Strange stock.

What is the significance of the orange cells J5 and K5?

[AIMAGIC]

Hi Myst,

I have sent you an optimised XDEV sheet for CELO. This high volatility play seems a perfect candidate for either XDEV or AIM.

No matter how I adjust the parameters when optimising the XDEV sheet, I can't seem to stop the positive values from appearing in the SELL MARKET ORDER column.

It seems to happen on days where the price passes through both the HIGH % DEV and the LOW % DEV bands. Is this something to do with the asset's volatility?

Would you leave the results as they are, with the positive values appearing from time to time and only optimise for the highest return over the look-back period?

I'd appreciate it if you could, perhaps, share the results after your optimisation.

Alex

[Myst]

Aimagic,

I have shared my Ethereum sheet with you. Once again it is very interesting to note that we achieved nearly identical ROI returns even though our parameters are very different.

My optimal minimum signal turned out to be 6%.

Compare the charts between sheets. Trying to figure out any conclusions based on the data.

[AIMAGIC]

Hi Myst,

I have sent you the ETH daily sheet.

Let me know what you think of the optimisation. Perhaps a more unusual outcome is the minimum signal % of 15%. I’d be interested to know if that is something you would ever trade.

Honestly, I taught my self. As I bump into things I want to improve I look it up on Google!

It is a privilege to look through your work and a joy to add to the work you have already done.

Alex

[Myst]

Awesome updates!

How is it you know excel (or google sheet) formulas so well? Do you work with them in your daily job?

Really appreciate your efforts....thanks much!

If you could put together an Ethereum daily sheet starting from 3/15/2021 that would be great!

thanks again,

[AIMAGIC]

Hi Myst,

I see what you mean. I may have a think about how it could be possible to make the MARKET ORDER correctly reflect the size of the CASH balance without messing up the SIGNAL column. It may not be possible.

I corrected the CASH % column and added a new STOCK % column. Previously, slightly annoyingly, the CASH % column would show in excess of 100%. By basing the calculation on the STOCK VALUE column I have made it more accurate. Now, the two columns correctly show STOCK and CASH proportions after taking into account any BUY or SELL.

I added an INDEX formula to the Total Return in cells J4 and K4 so you don’t manually have to input the cell in the last rows AC and AD each time after updating the spreadsheet. Hopefully this will make updating the spreadsheet even quicker.

You asked about the MDD calculation earlier. In each row, the formula subtracts the current price from the highest price in the rows above. If the price is beneath the last high, the formula tells you how many % drawdown you are in. If the price is above the last highest price then a “0” is entered into the cell. This way, on every row you see the % drawdown since the last high, both for XDEV and for B&H. The (pink) cells at the top of both columns show the Max Drawdown for XDEV and B&H over the look-back period. It is interesting to compared the performance in the chart below.

In cells J5 and K5 I added a CAGR calculation. This annualises the total return. Depending on what timeframe (Daily, Weekly) the numerator in the second half of the formula has to be either “365” or “52”.

I can save you some time, if you tell me what asset and from which date, and I'll send you a version of the spreadsheet with all the updates.

Best

Alex

[Myst]

No error

It doesn't matter what the Market Order says in those instances. The sheet formula will only calculate what is available in the cash value column to calculate the actual shares purchased.

You can do the math yourself to confirm if you like.

[AIMAGIC]

Hi Myst,

I’ve sent you an XDEV sheet of YINN.

I spotted what could be an error in the MARKET ORDER calculation.

For example in cells: V70, V109, V117.

In these cells the MARKET ORDER is greater than the CASH VALUE in the adjacent cell.

Unless I am mistaken it seems this would not be possible in practice.

Do we need to amend the formula so that in instances like this the MARKET ORDER is never greater than the CASH VALUE?

Kind regards,

Alex

[Myst]

I use my longer term X_DEV sheets and the X_DEV program as a guide for trading. I also check the aimagic sheets now as well. I use them as tools in combination with my technical analysis on the price charts. I look at a variety of time frames, I check the fundamental news sources, and even onchain and sentiment analysis. I've been doing this for so long now it is an actual lifestyle. I believe trading is simply too difficult an endeavor for anyone who isn't at least 90% dedicated to it everyday. Saying that, I developed the X_DEV program to help those that don't have the time or interest for 90% or better dedication. As a standalone program anyone with a modicum of interest can be successful at building wealth using it's features.

As stated in an earlier post I check the buy and sell targets on both X_DEV and AIMagic sheets. As the day progresses, since I have the luxury of watching the price action I decide if the targets are going to be exceeded or not. If not, I'll enter at the target price, if they look like they will be exceeded and my other analysis suggests they will be exceeded I will wait until I feel the entry is optimal.

For those that do not do this for a living, I'd simply check the targets before market open and place limit orders. Or if trading crypto, simply place the limit orders at any time. If your orders are close to the spreadsheet targets your ROI will not vary much at all. The goal is to come close, not to be exact.

[AIMAGIC]

Hi Myst,

Can you run me through your procedure for opening a trade when a BUY or SELL signal is generated after inputting the Close price into the XDEV or the AIM spreadsheet?

Do you ever enter limit orders at the CLOSE price and then hope the order is triggered the next day?

Perhaps, trading cryptocurrencies this is not so big a problem as the markets trade 24 hours.

However, for example, with MSTR, do you fin that the price at which you actually BUY and SELL is different to the CLOSE price that generated the SIGNALS shown in the spreadsheet?

If you make the trade the next day at the open, do you often BUY or SELL at different prices to those which are recorded as the CLOSE from the day before?

Will this mean that your results vary substantially to the results shown in the spreadsheet?

Is there any way to reflect these differences in your spreadsheet or, do you simply continue trying to follow the signals at slightly different prices each day?

Have you ever tried to take the signal just before the close each day instead?

Alex

[Myst]

The Graph - X_DEV Update:

[Myst]

I've updated and revised your WEEKLY BTC AIM sheet.

[Myst]

A buy or sell signal is initiated as soon as price has exceeded the upper (sell) or lower(buy) X_DEV line.

As long as the market order is above the minimum limit you have set the order is initiated. The number of consecutive buys or sells is irrelevant. You just follow the system signals. Sometimes you will go all in or to an all cash status but not always. Depends on price action.

[AIMStudent]

The power of scaling in & out...

Lots of groupings of consecutive buys over four and five days. Do you have any rules around number of back-to-back buys (and sells) or do you just buy until you run out of cash?

-AIMStudent

[AIMAGIC]

Fantastic Result

[Myst]

The power of scaling in near highs and scaling out near lows. MSTR is one of my current favorites and a large part of my IRA fund.

Note while the buy and holder lost -7.96% the X_DEV'r managed a +121.72% return simply by turning over shares and compounding the returns over and over. I have the cash line displayed on the chart so you can see how X_DEV went to all cash and all in several times. It didn't even have to nail the exact tops or bottoms for exceptional returns. Great play for X_DEV and AIM

[Myst]

The X_DEV software forecasts the price targets for each day, so it assumes the user placed limit orders at the target. So yes, it calculates the ROI based on the targets being executed precisely at the target order.

The X_DEV program also has a portfolio feature so users can enter their exact buys or sells in the event they did not enter on the exact target the program called for and the dollar amount they invested. This would calculate ROI and targets based on their inputs.

[AIMStudent]

Thanks Myst. Additional follow-up.

Thanks for clarifying.

An additional follow-up. How does the SW determine the buys and sells? Not the calculations, but how they are entered. For example, does it assume the price of the buy/sell forecast in the SW is the exact tick amount achieved on the day that price is encountered, similar to having a standing limit order? If so, is there any allowance for slippage? Or does it assume closing price on day the price is encountered. Or next session's open?

Thanks,
-AIMStudent

[Myst]

Aimstudent and Aimagic,

I was in error on an earlier post regarding the calculation of Buy and Hold % return and X_DEV % returns listed on the X_DEV software.

The calculation is NOT made from the starting date. It is made from the initial buy price.

Even though the start date on the chart was 2/2/2021 the initial buy was not until 2/23/21 at $761.5

(690.02/761.5)-1 = -9.39%


On the chart with the starting date of 5/2/2021 the initial buy was on 5/19/2021 at $422.86

(690.02/422.86)-1 = +63.18%

Apologies for the confusion!

[AIMStudent]

ST & LT on MSTR - B&H P/L?

Hey Myst,

Apparently I'm missing something in my understanding of your calculations for B&H P/L.

MSTR Close
02/02/21: $ 687.92
11/22/21: $ 690.02
B&H P/L: (690.02/687.92)-1 = 0.31%

05/03/21: $ 635.62 # 05/02/21 is a Sunday
B&H P/L: (690.02/635.62)-1 = 8.56%

Even if I convert to shares at the start (and have small cash left over from rounding shares down), use opening price, previous close, etc. the B&H % P/L is nowhere near what you are showing.

What am I missing...?

Thanks,
-AIMStudent

[Myst]

MST 50% starting cash. It may be that the safe bet would be to start with 50%. Need further study.

[Myst]

Examples of shorter and longer term time frames on MSTR with starting cash at 5% and 100%

[Myst]

There is no hard and fast rule as too how many months of data you need to backtest to get a good optimization. It's just important to get a minimum of 2 buy signals and 2 sell signals. Since you are using the spreadsheets instead of the program now, you should check your optimization against the charts, the 8sma and the dev factor to see if you are hitting good signals and you can't get the ROI any higher by any significant degree.

Then start a new play by using those settings going forward. It's probably best to be diligent in the early days of a new play for trend shifts, less so as the play matures. The starting cash is probably the least important metric in the setup. Even if you are off optimal initial starting cash the system will auto correct for that metric as the play matures. I haven't noticed major differences in ROI by altering starting cash by a large degree. On some plays it might but on average it doesn't seem to, especially as time goes by.

Checking the parameters for optimum ROI every 3 to six weeks should be all that is needed to keep performance near optimal.

[Myst]

So far, the optimizations look decent though I haven't analyzed everything to completion. I will let you know when I have and if I've found any "mistakes".

The X_DEV program does show profit and loss based on date you started (for the Buy and Hold profit/loss) and the day of your first buy (X_DEV profit/loss). Using the program this is usually the same day. The spreadsheet may differ.

Since I have the benefit of using the X_DEV program I will start by grabbing at least 6 months of data and then finding a significant low for a starting point, but at least having 2 buys and 2 sells for a decent backtest study. You are trying to find an average trading range to wrap the bands around. If a play has had a huge trend shift recently (like MANA for example), you won't be able to get a good representative sample for an X_DEV study and will have to wait a few weeks for the range to develop.

Without the benefit of the program I would pull up a chart on tradingview and look for a decent ranging period over the course of the past 6 months to a year. Find a low in range and start the backtest study from there.

BTW, it is very interesting to see that though we used different parameters for MSTR our ending ROI's are nearly identical. Speaks to the robustness of the system.

[Myst]

Your spreadsheet skills are vastly superior to mine.....I'm still very much a rookie. I'm trying to wrap my head around your MAX drawdown reasoning and starting cash calculations. The starting CASH formula seems great!

Perhaps you can expand a bit on the formula's for the Drawdown columns?

[Myst]

Aimagic,

Yes, X_DEV is a tool created to capitalize on short term trades over longer term time frames to compound returns. This is how I've always traded and it has been a successful strategy. I developed it after months of trying to tweak AIM to fit my philosophy and trading style. This is why AIM and X_DEV share some similarity.


You can start any play at any time. I think the Zones I came up with will serve as a decent starting metric for initial cash to allocate to the play. After about 20-30 days the user can optimize the spreadsheet and it should sort out all the metrics to the point of "robustness" going forward.

I rarely trade more than 5 assets at a time so I have sheets for all of them and keep them updated daily. It only takes a few minutes.

You have caught on very quickly so no worries about messing up the optimizations. Your MATIC sheet looks great!

[Myst]

Aimagic,

I use a combination of "eyeballing" the X_DEV bands (in relation to the high and low bands) and the charts on the bottom of the sheet to determine significant trend changes. The directional indicator is great for immediate bullish or bearish trend by viewing whether the green +DMI or the red -DMI is on top. Green bullish, red bearish. The blue ADX line indicates the strength of the trend. Rising ADX stronger trend, falling ADX weaker trend. The RSI, CCI, Williams %R, and DEV factor can determine how oversold or overbought current price action is. The strongest signals are when bullish or bearish divergences occur on these indicators.

I usually won't make a decision to change an X_DEV band parameter until at least 2-4 weeks of price action to confirm the trend change. But as stated before, if price has exceeded your last sell for 3 days, or if price is lower than your last buy for 3 days it is a strong signal that a significant trend reversal has occurred so a band parameter change may be necessary soon.

The $150 is a personal choice. I don't like to make extremely small dollar amount trades and would rather wait till a significant signal is given in the form of a larger dollar amount. This amount can be anything you prefer, and will lead to some ROI change depending on how small or large the amount you decide on. You will note on the X_DEV program there are parameters for minimum buy and sell amounts. This is why I included that feature in the sheet.

[AIMAGIC]

Hi Myst,

General questions I have as a result of modelling your latest updates…

How many months data do I need to start?

Does your experience suggest it is better either to:

Take 3 months of data before the start date

Or, as per the XDEV manual:

Start a new play and look for 2 BUY and 2 SELL signals before considering the spreadsheet optimised.

If you take the LINK chart, for instance. The play was started 7 7 21. By optimising from today (22 11 21) to that start date suggests the optimal starting CASH would have been 66.67%.

However, using data provided in the 3 months before 7 7 21, would have have had you 30% in CASH.

Do these anomalies get "ironed out" by re-optimising over time?

Leaving the settings unchanged going forward would probably have meant missing out on many BUY and SELL signals.

Is the answer, therefore, after a big drop, to wait 3 or 4 weeks and re-optimise?

By re-optimising, do we capture more of the trend?

How quickly does XDEV pick up a new trend?

Do we get the benefit of any recent price information by re-optimising every 3-4 weeks?

I think the spreadsheet doesn’t accurately show the results we’d get trading since the spreadsheet cannot show the result of these ongoing optimisations.

All the best,

Alex

[AIMAGIC]

Hey Myst,

I have sent you the X_DEV sheet of this play. Let let me know if I have made any obvious mistakes.

Since this play was started later and avoided the May / June drop in Cryptocurrency prices it has had less chance to pull ahead of B&H. I was able to generate a return of 76% for X_DEV and 43.4% or B&H.

Alex

[AIMAGIC]

Hi Myst,

MATIC seems to have been the least successful play. On a risk reward basis, not a great play. Perhaps this is due to the almost 50% drop in price soon after the 10 6 21 entry.

Even so, X_DEV is up 52% versus 19% for B&H.

Of interest in the optimisation was that X_DEV made 4 SELLS (22 & 23 & 26 July & 7 August) at a price below the previous buy (8 7 21), which I have not seen on any other sheet before.

This optimisation made me think that timing the BUYS and SELLS to only take a trade when CASH levels indicated are at 0% (all in) or 100% (all out) which might allow one to take fewer but higher return trades.

Alex

[AIMAGIC]

Hi Myst,

Here, I was able to generate an X_DEV return of 89% versus 13% B&H.

The "anomaly" in this optimisation was entering a negative (or -) percent for the high dev band, which produced the best result.

Alex

[AIMAGIC]

Hi Myst,

I have replicated each of the cryptocurrencies and the one stock for which you posted an update.

I am sharing them with you, perhaps you’ll let me know if you spot any really serious optimisation mistakes : )

Generally, the Profit/Loss and B&H Profit/Loss shown by the X_DEV software seems to show different returns to the spreadsheets modelled over the periods indicated by the Start Date and End Date.

It looks like the software may calculate the returns from the date of the first X_DEV trade (?) not the period indicated by the start date and end date.

First up, The Graph….

Have you got any “rules” for when you like to throw up a chart?

Why start 11 5 21?

Would you now trade this play on the basis of the 6 months of data in the set up?

I was able to generate a 81.6% return with my optimisation.

After such a “badly timed” entry and an immediate 36% drop, X_DEV is now up 81%, while B&H is still down -36%.

Alex

[AIMAGIC]

Hi Myst,

I have shared a MSTR sheet.

I added two columns to calculate current drawdown and, then, Maximum Drawdown for XDEV and for STOCK over the look-back period. There is also a chart comparing them.

Calculating the Maximum Drawdown allows us to divide Profit by Maximum Drawdown over the look-back period to give an indication of Reward to Risk.

This makes clear XDEV’s outperformance on a risk-adjusted basis versus B&H, may help in optimisation, or may allow us to make risk-adjusted comparisons between optimisations and assets.

Let us see how useful it is.

I altered the START CASH % formula slightly so that START CASH % indicated now incorporates any BUY or SELL MARKET ORDER that same period.

The START CASH % is now an accurate reflection of what the START CASH % should be that period after accounting for any BUY or SELL signals XDEV is indicating.

Best regards,

Alex

[AIMAGIC]

Hey Myst,

Thank you very much for you thorough and comprehensive response and for revising the CARA weekly sample.

"X_DEV is a tool meant to be used on high volatility plays optimised for a 3-6 month time frame. I created it this way to suit my personal trading style and philosophy."

I can see that running 7 years of data through the X_DEV algorithm does not make sense. I am beginning to see the way to use XDEV is to "throw up a chart" using 3 months of data and start trading.

Have you found some way of indicating and timing when you start a new play? Do you start a new play at an asset's lows"? Perhaps the low of a 52-week range? Or near the 200 DMA?

Once you have begun a play, how regularly do you find yourself starting a new sheet?

I'll keep modelling various assets, hopefully I'll learn to make less of a mess of the optimisation!

Kind regards,

Alex