|
PureBytes Links
Trading Reference Links
|
Walter Lake wrote:
>
> Hi Ton
>
> Here's chapter 3 - page 25+
>
> Lane's Stochastic
> Unsmoothed (fast) stochastic
> k=100*(close-LL(q)) / (HH(q)-LL(q))
> q = {lookback period}
>
> HH(q) = highest high in last q bars
> LL(q) = lowest low in last q bars
> one form of slow stochastic = 3 day SMA smoothing
> d% = 100*(SMA(3) of (close-LL(q)) / (SMA(3) of (HH(q)-LL(q))
>
> double (EMA) smoothed stochastic
> DS(q,r,s) = 100*(EMA(EMA(close-LL(q),r),s)) / (EMA(EMA(HH(q)-LL(q),r),s))
>
> Stochastic Momentum = SM {distance of close - midpoint}
> SM(q) = close - 0.5 * (HH(q) + LL(q))
>
> Stochastic Momentum Index = SMI
> SMI = 100*(EMA(EMA(SM(q),r),s)) / (0.5*EMA(EMA(HH(q)-LL(q),r),s))
> threshholds set at +100 & -100
>
> basic configuration
> q {lookback} = 13 days
> EMA smoothing of 25 days
>
> Signal line is the EMA of SMI(q,r,s)
> set in the range of 3 to 12 bars
>
> Double smoothing of SMI(q,r,s) is obtained by using any two of q,r,s
> parameters while holding the remaining parameter fixed.
>
> Stochastic Momentum Index = SMI(13,25,1)
> plus signal line
> thresholds set at +40 & -40
>
> cumulative sum of Stoachastic Momentum with q = 20 day lookback is similiar
> to 20 day EMA of close but ineresting
>
> Stochastic momentum for large intervals = SMI(300,5,1)
> first smoothing for q = 300
> second EMA smoothing of 5 days
> only lag was due to r day EMA
>
> shorter time periods produce divergences
> compare 20 day stochastic momentum with 5 day smoothing
> with stochastic momentum index SMI(20,5,1)
> disparity is due to 5 day EMA for SMI
> Vs 3 day SMA for slow stochastic
>
> EMA smoothing can be increased from 5 to 20 days
> SMI(20,20,1)
> trends as prices trend
>
> SMI can be increased to SMI(20,60,1) to define trend in prices
> with slow stochastic used as entry (exit) vehicle
>
> Long trend SMI = SMI(20,150,1)
> corrects any anomalities by increasing r from 20 to 150
> smooth and with very little lag
>
> - 0 -
>
> will continue later with rest of chapter
> 2 day and 1 day stochastics
>
> Best regards
>
> Walter
Thanks very much Walter for taking the time and patience to do this
for those of us who really appreciate it a great deal.
Randy
|