Category Archives: Uncategorized

Macintosh version of Excel System Back-Tester Available

It was recently brought to my attention that the Excel System Back-Tester(ESB) did not function properly on the MAC OS X.  In other words it bombed when trying to open a comma delimited data file and also when one tried to run an algorithm.  Thanks to a purchaser of the UATSTB I was able to fix the bugs without removing any functionality.  I will post the Macintosh version here as well as have WILEY put it on the book’s website.  Sorry for any inconvenience this may have caused.  Here is the the link:

 

ESB Macintosh

Using Quandl Data

If you haven’t come across the great data resource Quandl.com I highly suggest in doing so.  A  portion of the data that I used in writing my latest book came from Quandl, specifically the wiki futures or CHRIS database.  Here is the link:

https://www.quandl.com/data/CHRIS

There is a ton of free futures data.  The different contracts are concatenated into large files.  However, the rollover discount is not taken into consideration so the data “as-is” is somewhat un – testable.  I am in the process of scrubbing the data as well as creating  a “Panama” adjustment to the contracts in the large files.  As soon as I complete this task I will provide the data on this website.  Purchasers of my latest book will find 10+ plus years of history of continuous futures data that I pieced together from several different sources, including Quandl.

Further Clarification on Data Aliasing

I was speaking with Mike Chalek on the phone this weekend concerning Data Aliasing and he felt this post was a little confusing. After re-reading it I can see where he is coming from. Using the same example let me see if I can clarify: assume the trading day is Wednesday and you want to keep track of the slope of a 19-day weighted moving average of data2 (weekly bars) by using a variable. The following code will give an erroneous result:

wAvg = wAverage(c of data2,19);
mySlope = wAvg – wAvg[1];

If you interrogate mySlope intra-week then it will always be equal to zero. The wAvg is by default tied to data1 which in this case is daily bars. So the value of wAvg is carried over from one day to the next. It only changes when the average of the weekly bar changes and that only occurs on Friday.

There are two possible solutions:

Without the use of data aliasing – inLine function calls
mySlope = wAverage(c of data2,19) – wAverage(c[1] of data2,19) ;

With the use of data aliasing –
vars: wAvg(close of data2,0);

wAvg = wAverage(c of data2,19);
mySlop = wAvg – wAvg[1];

Either examples will work, but if you have several variables tied to a different data stream, then the code will be much cleaner looking using data aliasing – plus it cuts down on multiple function calls.

Using Multiple Time Frames in a Strategy

I have been working on a project where the strategy combined daily and weekly bars.  Keeping track of the two time frames was, at one time, not that easy.  However, with TradeStation’s Data Aliasing it is no problem at all.  We all know that Data 1 is the highest resolution time frame and is the one used for trade execution.   Data 2 can be a different market or a different time from of the same market.  TradeStation allows for multiple data streams.  Take a look at the following output in table 1.  Wavg is a nine period moving average of weekly crude data.  Wavg[1] is the prior value of the moving average.  If you wanted to make a trading decision on a daily bar basis by looking at the slope of the Wavg you couldn’t.  The Wavg and Wavg[1] only changes at the beginning of the next week.  Most traders want to be able to make a trading decision intra-week by examining the current values of the Davg1, Davg2 and the slope of Wavg.  During the week the slope of Wavg is ZERO.

table 1
Date    Davg1 Davg2 Wavg Wavg[1]
1151019 46.94 46.38 46.17 46.17
1151020 47.01 46.54 46.17 46.17
1151021 47.00 46.69 46.17 46.17
1151022 46.95 46.74 46.17 46.17
1151023 46.93 46.70 46.54 46.17<< changed here
1151026 46.83 46.55 46.54 46.54
1151027 46.71 46.47 46.54 46.54
1151028 46.74 46.44 46.54 46.54
1151029 46.74 46.40 46.54 46.54
1151030 46.73 46.39 46.60 46.54
1151102 46.57 46.37 46.60 46.60
1151103 46.55 46.45 46.60 46.60
1151104 46.36 46.44 46.60 46.60

Now look at table 2.   The Wavg is not being updated on a daily  basis but on a weekly basis.  The current Wavg doesn’t become the prior Wavg on each daily bar.  Wavg[1] stays the same until a new weekly bar occurs.  You can now make a trading decision intra-week by examining the slope of the Wavg.  Each time frame update should only occur when a new bar of that same time frame is generated.  This feature is really cool and is easy to implement.  

table2
Date      Davg1 Davg2 Wavg Wavg[1]
1151019 46.94 46.38 46.17 45.75 < notice how the Wavg and Wavg[1] are always different
1151020 47.01 46.54 46.17 45.75
1151021 47.00 46.69 46.17 45.75
1151022 46.95 46.74 46.17 45.75
1151023 46.93 46.70 46.54 46.17
1151026 46.83 46.55 46.54 46.17
1151027 46.71 46.47 46.54 46.17
1151028 46.74 46.44 46.54 46.17
1151029 46.74 46.40 46.54 46.17
1151030 46.73 46.39 46.60 46.54
1151102 46.57 46.37 46.60 46.54
1151103 46.55 46.45 46.60 46.54
1151104 46.36 46.44 46.60 46.54

 

Here is the code that utilizes Data Aliasing. All I did was declare the weekly avg variable and tied it to data2.

vars: mavShortDaily(0),mavLongDaily(0);
vars: mavWeekly(0,data2);

mavShortDaily = average(c,19);
mavLongDaily = average(c,39);

mavWeekly = average(C of data2, 9);

If mavShortDaily > mavLongDaily and mavWeekly > mavWeekly[1] then buy this bar on close;
If mavShortDaily < mavLongDaily and mavWeekly < mavWeekly[1] then sellshort this bar on close;

print(date," ",mavShortDaily," ",mavLongDaily," ",mavWeekly," ",mavWeekly[1]);

Notice how the variable mavWeekly was tied to data2. When you delcare a variable that is tied to another data other than data1 you can put the data stream right in the variable delcaration : mavWeekly(0,data2).

How to Round Up/Down To Nearest Tick in EasyLanguage

This is how you round to the nearest tick in EasyLanguage – helpful when plotting
price based indicators. Also the formula for calculating the min tick value is given.



vars: minTick(0),testPrice(0);

minTick = minMove/priceScale;
testPrice = close * .21 * range;

// round up
value1 = testPrice + (minTick-mod(testPrice,minTick));
// round dn
value2 = testPrice - (mod(testPrice,minTick));

{mod is a call to the modulus function
 aka remainder function -- mod(12,5) = 2 -- 12/5 = 2 Remainder 2
 say ES testPrice = 1123.57
     minTick = .25
     1123.57 + (0.25 - mod(1123.57,0.25)) = 1123.57 + 0.25 - 0.07 = 1123.75}

Correction to Thermostat and Bandit Description in Book

A very astute reader of the BWTSwTS2 has brought to my attention  errors in my description of the Thermostat and Bollinger Bandit algorithms. In the Thermo description I incorrectly used the words yesterday and today. The code is correct in the book. Thanks to John for finding this!

Corrected description follows:

….If today’s closing price is greater than the average of today’s high,low and close, then we feel tomorrow’s action will probably be bearish. However, if today’s closing price is less than or equal to the average of today’s high, low, and close, then tomorrow’s market will behave in a bullish manner.

In addition John uncovered a typo as well for the Bollinger Bandit description – when I stated BELOW I meant ABOVE and vice versa.

Corrected description follows:

If liqPoint is BELOW the upband, we will liquidate a long position if today’s market action  <= liqPoint.

 If liqPoint is ABOVE the dnband, we will liquidate a long position if today’s market action  >= liqPoint.

 

Correct King Keltner Easy Language

Several have brought it to my attention that the King Keltner code in the book is missing a couple of lines. Here’s the complete code in its entirety. Thanks for bringing this to my attention.



[LegacyColorValue = true]; 

{King Keltner Program
King Keltner by George Pruitt -- based on trading system presented by Chester Keltner
 -- an example of a simple, robust and effective strategy}

Inputs: avgLength(40),atrLength(40);
Vars: upBand(0),dnBand(0),liquidPoint(0),movAvgVal(0);

movAvgVal = average((h+l+c)/3,avgLength);

upBand = movAvgVal + AvgTrueRange(atrLength);
dnBand = movAvgVal - AvgTrueRange(atrLength);

{Remember buy stops are above the market and sell stops are below the market
 -- if the market gaps above the buy stop, then the order turns into a market order
 vice versa for the sell stop}

if(movAvgVal > movAvgVal[1]) then Buy ("KKBuy") tomorrow at upBand stop;
if(movAvgVal < movAvgVal[1]) then SellShort("KKSell")tomorrow at dnBand stop;

liquidPoint = movAvgVal;
 
if(MarketPosition = 1) then Sell tomorrow at liquidPoint stop;
if(MarketPosition =-1) then BuyTocover tomorrow at liquidPoint stop;