and turtles still work?
MattW,
I think that quote should be embedded on every traders brain. You are quite correct and how many traders have systems, especially new ones, that they continually tinker, let emotions get involved etc, and never truly allow the positive expectancy come to light.
Reading your quote, you have articulated exactly what people can have control with their trading but rarely put alot of time into, as they look for the perfect entry.
A win ratio of 37.5% - 40% with 1:2RR is profitable especially compounding profits. If new traders focussed on achieving these numbers at a minimum they can become successful. Then if they feel they need to search for a holy grail at least they have one system working for them.
LMAO! This is [U]EXACTLY[/U] the journey Iâm on now! Sometimes I forget that others are attempting (and some succeeding at) this gameâŚ
Iâm at the back of a year of break even trading across the g/u, e/u & e/j with a strict 2-in-1-out model. Just over 40% WR across all three pairs. Should have dug deeper a lot sooner but, eventually, looked into the MFE of my trades and the effect of slippage & commissionsâŚwhich vary significantly between pairs (along with the WR)
After alot of thought, moved from IB (slippage on exits on losers ranged from 0.1R to a hefty 0.3R) to IG index and reduced trading activity to e/u only. Backtesting would suggest that the wider spread with IG will lower the ~48% WR by several % BUT this is far outweighed by:
- Savings on commissions
- Zero slippage on losers
In short, keeping more of the 0.27 Expectancy.
As Shr1k points out though, variance can make it difficult to trade through the inevitable draw downs that come. 10 % isnât unusual and 8-in-row is a more frequent event than Iâd like it to be!
MM
@ Smithy⌠Indeed you are correct⌠I trade with a 2:1 Win:Los ratio and 40% of my trades are succesful, I made 123% on my principal last yearâŚThe most loss I had was in december when I lost 8 days in a rowâŚI started trading real money in July of last yearâŚCan you tell me how to get a number in a spreadsheet rounded off to the nearest 10, for example - IF THE VALUE IN A CELL IS RETURN TO 12,454 - HOW DP I GET IT TO AUTOMATICALLY ROUND OFF TO 12,450??
Wow, thatâs pretty impressive man.
Thanks MattâŚ
What are these turtles that you speak of?
anyone run a random entry back test (or forward test) with a reward:risk of 3:1 as opposed to the discussed 2:1? This way your winners are 50% larger (3 as opposed to 2) plus I cant imagine your win rate dropping to more than 28-30% winners.
The larger win size will also make slippage less of a factor
A very brief Turtle history:
In 1983, Richard Dennis and William Eckhardt, two Chicago commodities traders, disagreed over the question: are expert traders born or made? Eckhardt said that whether or not a person could become an expert trader was a matter of genetics; but, Dennis said he could teach ANYONE to be an expert trader IF they would follow his rules to the letter.
So, Dennis and Eckhardt set up the Turtle Trader Experiment, recruited 13 newbies (who were called the Turtles), and set about to test Dennisâ claim that he could teach these guys high-level trading. The Turtles were taught Dennisâ trading rules (now generally referred to as The Turtle Trading Rules), and they were contractually bound to keep the rules secret for 10 years after the experiment ended.
A second generation of Turtles were trained, as well; and, to avoid confusion, the original 13 became known as The Original Turtles.
Years later, a writer named Michael Covel started writing about the Dennis/Eckhardt experiment, about the Original Turtles, about the second-generation Turtles, and most importantly about The Turtle Trading Rules. Covel has become, not only the âbiographerâ of the Turtles, but also a trading guru teaching âtrend-followingâ techniques.
One of the Original Turtles, a guy named Curtis Faith, then came forward and said basically that Michael Covel was a scam because he had only second-hand information about the Turtle experiment and its participants, because he got some things wrong concerning the Rules, and because he was a guy who didnât even trade. Curtis Faithâs organization, OriginalTurtles.org â whose motto was âFighting the scams, frauds and charlatansâ â published a short book titled The Original Turtle Trading Rules, in order to set the record straight. Faith claims that his book is the only source for the true Turtle Trading Rules, as they were taught by Dennis and Eckhardt.
Curtis Faith seems to have had a troubled life, professionally at least, in the years since writing The Original Turtle Trading Rules.
In recent years, Michael Covelâs website has documented the downward spiral of Curtis Faith, even going so far as to call him âimpairedâ â Curtis Faith: Firm Permanently Barred by U.S. Government | TurtleTrader Obviously, there is âbad bloodâ between these two.
In one final irony, the Curtis Faith website OriginalTurtles.org, which referred to Covel (not by name) as a scam, now appears to be controlled by Covel. The website name has been changed slightly, but if you try to go to OriginalTurtles.org, you will end up here â The Official Site of the Original Turtles â Same logo, new name, new website, now proudly selling Michael Covelâs books. Go figure.
I have attached a .pdf of Curtis Faithâs The Original Turtle Trading Rules, written in 2003 (before many of Faithâs problems had surfaced). Itâs worth a read.
turtlerules.pdf (271 KB)
1 Like
interesting readâŚthanks for sharing.
in regards to the Turtles⌠Turtle Soup comes to mind hahaha
trade how the banks trade, and keep your expectations low and your focus on consistency and I donât think you will go too wrongâŚ
Going back to original commentaryâŚfor those who wanted to try themselves in excelâŚmake sure the size of the trade is rounded (i.e you canât trade a 28.452k lot size).
The short answer, as many have pointed out, is yesâŚprofitability with those numbers is possible. But itâs a foolish question. Too many people ask if they can be profitable with xyz. You need to really determine the reliability of profit and risk of ruin.
Some notes on those excel stipulations (given a 50/-25 P/L). After running a few dozen times, the average pips per trade ranged from about 0 to 6. The average of the avg pips per trade seemed to be about 2-3. The actual math says the avg should be 3.125 pips.
If spread is NOT taken into account (and depending on the broker) could turn this into a losing âsystemâ on average. Including likely slippage from time to time can even further reduce profitability and in combination with spread can again turn it into a losing system on average. Outside of that obvious math, non-adherence to the rules (happens with everyone at some point) can further cause harm and turn the system negative.
Just remember, there are no right answers to wrong questions. 
This is an old thread, which has just recently been resurrected.
The original poster hasnât rejoined the discussion, so weâre proceeding without him. Maybe heâll jump back in.
In the meantime, I wanted to comment on âstrings of lossesâ, an issue just barely touched on, up to this point.
From the original poster:
I placed the last part of smithyâs quote in bold type for emphasis.
Thereâs a nifty equation which predicts the average recurrence of strings of losses of various lengths, based on one simple parameter: the W/L ratio. Iâll go through the math at the end of this post, for anyone whoâs interested.
The string of losses that smithy mentioned above (8 losses in a row) will occur, on average every 11 days, based on the trading plan outlined in his first post (W/L = 37.5/62.5, 4 trades per day, 20 trades per week, etc.). He suggested that many traders might panic after 8 consecutive losses.
What about even longer strings of losses? Hereâs what the equation predicts, as an overall average, for smithyâs trading plan:
⢠a string of 6 losses will occur, on average, every 5 days
⢠a string of 8 losses â every 11 days
⢠a string of 10 losses â every 28 days
⢠a string of 12 losses â every 71 days
⢠a string of 14 losses â every 181 days
âDaysâ above means trading days (5 trading days per week). I have rounded the numbers up to the nearest whole day.
With 25 pips down the toilet for each loss, how many traders can withstand 10, or 12, or 14 losses in a row, without throwing in the towel, or starting to bend their rules?
If the W/L ratio were improved to 40/60, a string of 10 losses would occur, on average, once every 42 days.
And, if the W/L ratio were improved to 50/50, a string of 10 losses would occur, on average, only once every 256 days (vs. 28 days for smithyâs trading plan).
These numbers will prove correct over the very long term. In the short term, anything is possible. With a W/L ratio of 37.5/62.5 (smithyâs W/L), it might happen that months go by without a string of 10 or more losses. Or, 10 losses could occur 3 times in a month. Funny things, averages.
Hereâs the math behind these numbers.
We define the following variables:
W = number of wins per 100 trades
L = number of losses per 100 trades
a = 100/L
r = length of string (number of losses in a row)
N = expected number of trades required to produce a string of length r
Equation: N = a (exp r)
Sorry for the awkward notation. In the equation above, âexp râ means that r is an exponent. So, if r = 3, for example, the equation would read: N = aÂł. (I havenât been able to locate a keyboard symbol for variable exponents.)
In the discussion of smithyâs trading parameters, I said that a string of 10 losses would occur, on average, every 28 days. Hereâs how that was figured:
W = 37.5 wins per 100 trades
L = 62.5 losses per 100 trades
a = 100/62.5 = 1.6
r = 10 losses in a row
N = 1.6 (exp 10) = 109.95 trades = 27.49 days (at 4 trades per day) = 28 days (rounded up).
Hi Clint,
Very interesting figures, and I will investigate further. The four pairs I trade vary from 39.5 to 41% at this point after 170 days of trading. All either 50 pip win or 25 loss.
The timeframes are EURUSD 5pm & 6pm QLD time which is Frankfurt UK open, and GBPUSD 7pm & 8pm, 1 to 2 hour into UK session.
On those timeframes
Eurusd 5pm - Max in row losses - 8 (5 losses in a row 4 times)
Eurusd 6pm - Max in row losses - 6 (5 losses in a row 5 times)
Gbpusd 7pm - Max in row losses - 9 (5 losses in a row 2 times, 6 in a row once)
Gbpusd 8pm - Max in row losses - 15 (7 losses in a row once)
All the rest of the time is either 4 or less losses in a row.
When I discussed the 8 losses in a row I was talking sideways, Iâm not sure if that was clear, eg 26/8/10 & 27/8/10, I lost all 4 pairs on those two days. That has occurred 4 times in the 170 trading days, and twice with 3 days in a row -12 losses.
My trading is is purely looking at the statistics of a basic system to keep at a minimum 37.50% with 1:2 risk/reward. It is always good to read interesting posts to allow me to look further at possible flaws of what has been good so far, so thak you for your post.
Cheers
Does anyone know what happened to the turtles in the end?
Interesting stuff. Trading a decent strategy has often been compared to tossing a weighted coin. You know that overall the coin will come up head, say, 75% of the time, so you would happily place money on that outcome over time. However, you have no way to predict the result of an individual toss, or even the next 50 tosses. So the longer-term you view it, the more reliably you can expect to hit that 75%.
Getting to the point of not worrying about losing days (or weeks!) is a key battle when learning to trade.
Richard Dennis talks about it in his interview for Market Wizards (the original one), by Jack Schwager. An interesting read and worth getting if you have not read it. Confess I am a bit rusty and canât recall the detail, but from memory some of them drifted away and sort of lost interest, but the majority continued more seriously. Several of them made a lot of money, I think, but someone here might be able to prove my memory unreliableâŚ?! The majority (80% success rate from memory) did very well, proving Dennisâ contention. Not sure whether they still trade, the actual programmes were pretty short.
One of them was bronzed, and now sits atop a pedestal
in front of the library at the University of Maryland.
