|
PureBytes Links
Trading Reference Links
|
Hi there,
re: Rina Index and K-Ratio
from Rina / Omega help
>>RINA Index
>>
>>This proprietary index combines select net profit, time in the
>>market, and drawdown calculations into a single reward/risk ratio.
>>The larger the number the more efficient the strategy. Look for a
>>strategy with an index of 30 or more.
>>RINA Index = (Select Net Profit)/((Average Drawdown) x (Percent time
>>in the market))
>>
>>{ ** © 1987, 1999 Omega Research, Inc. ** }
I feel Rina index is too difficult to estimate
if you have limited set of information on
the investment product / strategy /portfolio.
This is especially the case for non-proprietary / third party
investment products.
As a matter of fact one day you will want to compare
some in-house developed strategies to different benchmarks.
I do not think Rina or K-ratio will give you much info
on this. The dissadvantage of Rina-index is it's only
standard to Rina and Omega products.
I will agree to Bob Fulks and many others' choice -- lovely
one number performance measure -- Sharpe ratio. You probably
should give more attention to this one.
May be one thing I really want to add is that I would use
a little bit modified "Sharpe Ratio 2" as per managedfutures.com/
Stark report users' guide terminology:
****** Start of quotation *************************************
14. Risk/Reward Ratios – A) Monthly Win/Loss Ratio (W/L) Number of
winning and losing months over the time period charted is shown along
with the average performance for each. A ratio is then determined for
a) number of winning months: number of losing months; b) winning month
average: losing month average. B) Return Retracement Ratio (RRR)
Measures return to risk. Return, in the numerator, is defined as the
trading program's annualized rate of return. The average maximum
retracement, in the denominator, is used to measure risk. C) ITR Ratio
(ITR) Measures return to risk. Return, in the numerator, is defined as
the trading program’s annualized rate of return. The worst drawdown
+.01, in the denominator, is used to measure risk. – includes
measurements of the amount of return from an investment per unit of
risk (volatility). A high ratio is desirable. D) Sharpe Ratio (Shrp);
Measures return to risk. Return, in the numerator, is defined as the
trading program’s annualized rate of return. Annualized monthly
standard deviation in the denominator is used to measure risk. E)
Sharpe Ratio II (Shrp2); Similar to Sharpe ratio (see above) but the
measure of risk used in the denominator is the trading program's
annualized monthly Semi Deviation. F) Sterling Ratio: Equals the
average annual compounded return for the past three-years/average
Maximum Annual Drawdown (MAD) +10%. The value of MAD is designated
positive while the average annual compounded return can be positive or
negative. (If this value is negative, however, the whole ratio is zero
since negative Sterling Ratios have no significant meaning).
****** End of quotation *************************************
I think the idea behind is if you are medium-term mentality trader
you will appreciate the turtles risk-adjusted performance. But original
Sharpe ratio penalizes the equity volatility caused by rare run-ups in
equity produced by all long-term trend-followers. BUT IT'S POSITIVE
EQUITY VOLATILITY. This is why the trading program's annualized monthly
Semi Deviation measure of risk used in the denominator of Sharpe ratio
2 looks much more informative personally for me.
But again it is not a standard and wide used. And you will have to
calculate it yourself every time you decide to compare the performance
of the different strategies, investment products and benchmarks.
re: K-ratio details
Unfortunately I have no original article from Futures Magazine
cause its electronic version is not available in archives
http://www.futuresmag.com/library/jan96/jan96contents.html
Rina / Omega help gives very little information
>> K-ratio
>>
>>This ratio is similar to the Sharpe Ratio. It differs in that it uses
>>linear regression techniques to measure the consistency of results
>>rough time. The higher the ratio, the greater the return in relation
to risk.
>>te This value will not display for strategies applied to intraday
charts.
>>
>>** © 1987, 1999 Omega Research, Inc. ** }
but let me refer you to post by Andrew Peskin <andrew@xxxxxxxxx>
to some of the forums on the K-ratio subject (I hope Andrew wont
mind ...)
************ Start of the the quotation *********************
Date: Mon, 11 Sep 2000 09:56:11 -0400
From: Andrew Peskin <andrew@xxxxxxxxx>
Reply-to: andrew@xxxxxxxxx
CC: systems-only@xxxxxxxxxxxxx
Subject: SO_System Performance Evaluation Metrics -- K-Ratio
Regarding the recent discussion of system performance evaluation
metrics, I
have found
that the K-Ratio is an excellent measure of system performance. This
ratio
measures both
profitability and consistency of returns and then returns them in one
single
number.
For a more detailed description of the logic behind the metric, please
refer
to the
article "Getting a Handle on True Performance" by Lars Kestner found in
the
January
1996 issue of Futures Magazine.
Below is BASIC code which calculates this ratio. It is a function
which
requires you
pass it an array of VAMI values and the number of periods in the array.
It
will
return the single K-Ratio Value.
<--------------- CODE --------------->
Public Function KRatio(VAMI() As Double, Periods As Integer) As Single
Dim lnVAMI() As Double
Dim Count As Integer
Dim SumX As Double
Dim SumY As Double
Dim SumXSqr As Double
Dim SumYSqr As Double
Dim SumXY As Double
Dim SSxx As Double
Dim SSxy As Double
Dim SSE As Double
Dim sSqr As Double
Dim YIntercept As Double
Dim Slope As Double
Dim SlopeStdError As Double
Dim YIntStdError As Double
Dim Result As Double
ReDim lnVAMI(1 To Periods)
For Count = 1 To Periods
If (VAMI(Count) > 0) Then
lnVAMI(Count) = Log(VAMI(Count))
Else
KRatio = -1
Exit Function
End If
Next Count
SumX = 0
SumXSqr = 0
SumY = 0
SumYSqr = 0
SumXY = 0
For Count = 1 To Periods
SumX = (SumX + Count)
SumXSqr = (SumXSqr + (Count ^ 2))
SumY = (SumY + lnVAMI(Count))
SumYSqr = (SumYSqr + (lnVAMI(Count) ^ 2))
SumXY = (SumXY + (Count * lnVAMI(Count)))
Next Count
SSxx = (SumXSqr - ((SumX ^ 2) / Periods))
SSxy = (SumXY - ((SumX * SumY) / Periods))
Slope = (SSxy / SSxx)
YIntercept = ((SumY / Periods) - (Slope * (SumX / Periods)))
SSE = ((SumYSqr) - (((SumY) ^ 2) / Periods) - (Slope * SSxy))
SlopeStdError = Sqr((SSE / (Periods - 2))) / Sqr(SSxx)
YIntStdError = Sqr(SlopeStdError)
Result = (Slope / (SlopeStdError * Sqr(Periods)))
KRatio = Result
End Function
***************** End of quotation ***************************
I hope this helps.
Best regards,
Valery Bartashevich
Minsk, Belarus
E-mail: bvs97f@xxxxxxxxx
ICQ#12107901
>> From: The Torch <iceman_nrg@xxxxxxxxx>
>>Subject: GEN: Rina Index and K-Ratio
>>
>>Anyone have an in-depth definition of what is the Rina
>>Index and K-Ratio and what is a good number to have?
>>
>>I am playing around with TS2000 and have some systems
>>showing a 300+ reading on the Rina index and dont know
>> that is good or bad.
P.S. Be carefull with this number (300+), man!
It looks curve-fit / too good to be true.
Just do some forward / out-of-sample testing.
May be it's Ok but just check it.
__________________________________________________
Do You Yahoo!?
Yahoo! Auctions - Buy the things you want at great prices.
http://auctions.yahoo.com/
------------------------ Yahoo! Groups Sponsor ---------------------~-~>
Make good on the promise you made at graduation to keep
in touch. Classmates.com has over 14 million registered
high school alumni--chances are you'll find your friends!
http://us.click.yahoo.com/l3joGB/DMUCAA/4ihDAA/zf_UlB/TM
---------------------------------------------------------------------_->
To unsubscribe from this group, send an email to:
realtraders-unsubscribe@xxxxxxxxxxxxxxx
Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
|