Another Failed Volatility Histeresis: Ehlers’s Own Idea

August 3, 2014

(This article was first published on QuantStrat TradeR » R, and kindly contributed to R-bloggers)

This week, I attempted to use Ehlers’s own idea from this presentation.

Essentially, the idea is that when an indicator is flat, line crossings can produce whipsaws, so add a fraction of the daily range to the lagged indicator, and see if the non-lagged indicator crosses the threshold. In this case, it’s an exponentially smoothed daily range that’s used to compute the bands. I ran this from 2012 through the present day at the time of this writing (July 14, 2014), as the original link goes through most of the 2000s. (Also, be sure you’re using my most up-to-date IKTrading package, as I updated the quandClean function to deal with some intraday messy data issues that had gone unnoticed before.)

The settings I used were John Ehlers’s original settings — that is, a 20 day analysis period, a 10 day exponential band smoothing (that is, the band is computed as .1*(high-low)+.9*band), entered upon the percent B (that is, the current FRAMA minus the low band over the difference of the bands), and the fraction is 1/10th of the daily range.

Here’s the indicator used:

FRAMAbands <- function(HLC, n=126, FC=1, SC=300, nBands=n/2, bandFrac=10, ...) {
  frama <- FRAMA(HLC, n=n, FC=FC, SC=SC, ...)
  band <- Hi(HLC) - Lo(HLC)
  band <- xts(filter(1/nBands*band, 1-1/nBands, method="recursive"),
  bandUp <- frama$trigger + band/bandFrac
  bandDn <- frama$trigger - band/bandFrac
  pctB <- (frama$FRAMA-bandDn)/(bandUp-bandDn)
  out <- cbind(frama, pctB)
  colnames(out) <- c("FRAMA", "trigger", "pctB")

And here’s the strategy code:


#trade sizing and initial equity settings
tradeSize <- 100000
initEq <- tradeSize*length(symbols) <- <- <- "FRAMA_BANDS_I"
initPortf(, symbols=symbols, initDate=initDate, currency='USD')
initAcct(,, initDate=initDate, currency='USD',initEq=initEq)
initOrders(, initDate=initDate)
strategy(, store=TRUE)

FC = 1
SC = 300
n = 20
triggerLag = 1
nBands = 10


add.indicator(, name="FRAMAbands",
              arguments=list(HLC=quote(HLC(mktdata)), FC=FC, SC=SC, 
                             n=n, triggerLag=triggerLag, nBands=nBands,

add.indicator(, name="lagATR", 
              arguments=list(HLC=quote(HLC(mktdata)), n=period), 

add.signal(, name="sigThreshold",
                          relationship="gt", cross=TRUE),

add.signal(, name="sigThreshold",
                          relationship="lt", cross=TRUE),

add.rule(, name="ruleSignal", 
         arguments=list(sigcol="longEntry", sigval=TRUE, ordertype="market", 
                        orderside="long", replace=FALSE, prefer="Open", osFUN=osDollarATR,
                        tradeSize=tradeSize, pctATR=pctATR, atrMod="X"), 
         type="enter", path.dep=TRUE)

add.rule(, name="ruleSignal", 
         arguments=list(sigcol="longExit", sigval=TRUE, orderqty="all", ordertype="market", 
                        orderside="long", replace=FALSE, prefer="Open"), 
         type="exit", path.dep=TRUE)

#apply strategy
t1 <- Sys.time()
out <- applyStrategy(,
t2 <- Sys.time()

#set up analytics
dateRange <- time(getPortfolio($summary)[-1]

Here are the results:

> (aggPF <- sum(tStats$Gross.Profits)/-sum(tStats$Gross.Losses))
[1] 0.956477
> (aggCorrect <- mean(tStats$Percent.Positive))
[1] 36.39737
> (numTrades <- sum(tStats$Num.Trades))
[1] 1778
> (meanAvgWLR <- mean(tStats$Avg.WinLoss.Ratio[tStats$Avg.WinLoss.Ratio < Inf], na.rm=TRUE))
[1] 1.678421

> print(t(durStats))
Min      1
Q1       2
Med      6
Mean     9
Q3      14
Max     65
WMin     1
WQ1      3
WMed    13
WMean   13
WQ3     19
WMax    65
LMin     1
LQ1      2
LMed     4
LMean    6
LQ3      8
LMax    57

[1] 0.08232023

> SharpeRatio.annualized(portfRets)
Annualized Sharpe Ratio (Rf=0%) -0.2476826
> Return.annualized(portfRets)
Annualized Return -0.03485231
> maxDrawdown(portfRets)
[1] 0.2632001

In short, it’s a loser over the past three years. Here’s the equity curve:

Now while it may have worked in the past (or something similar to it, using Ehlers’s filter indicator), it doesn’t seem to do so going forward.

I’ll leave this here for now as a demonstration of how to do Ehlers bands.

Thanks for reading.

To leave a comment for the author, please follow the link and comment on their blog: QuantStrat TradeR » R. offers daily e-mail updates about R news and tutorials on topics such as: Data science, Big Data, R jobs, visualization (ggplot2, Boxplots, maps, animation), programming (RStudio, Sweave, LaTeX, SQL, Eclipse, git, hadoop, Web Scraping) statistics (regression, PCA, time series, trading) and more...

If you got this far, why not subscribe for updates from the site? Choose your flavor: e-mail, twitter, RSS, or facebook...

Comments are closed.

Recent popular posts


Mango solutions

RStudio homepage

Zero Inflated Models and Generalized Linear Mixed Models with R

Quantide: statistical consulting and training


CRC R books series

Contact us if you wish to help support R-bloggers, and place your banner here.

Never miss an update!
Subscribe to R-bloggers to receive
e-mails with the latest R posts.
(You will not see this message again.)

Click here to close (This popup will not appear again)