Welcome to NexusFi: the best trading community on the planet, with over 150,000 members Sign Up Now for Free
Genuine reviews from real traders, not fake reviews from stealth vendors
Quality education from leading professional traders
We are a friendly, helpful, and positive community
We do not tolerate rude behavior, trolling, or vendors advertising in posts
We are here to help, just let us know what you need
You'll need to register in order to view the content of the threads and start contributing to our community. It's free for basic access, or support us by becoming an Elite Member -- see if you qualify for a discount below.
-- Big Mike, Site Administrator
(If you already have an account, login at the top of the page)
A Random Walk Stopped: My Journey to find the balance between Stoploss/ProfitTarget
This journal will explore the concept of an efficient but random market, and try to ascertain what stop placement will provide for the most profit in that market.
---Of course, the obvious theory behind the title is the Random Walk Hypothesis made popular in the 1973 book, A Random Walk Down Wall Street, by Burton Malkiel, but dating back as far a 1863. There is a plethera of information on the wikipedia entry here: Random walk hypothesis - Wikipedia, the free encyclopedia if you haven't read the book. More information on the mathmatical theory behind it is found here: Random walk - Wikipedia, the free encyclopedia.
---Essentially, the economic theory states that the markets are not able to be predicted, since the likelyhood of the price going in one direction at any moment is equal to the likelyhood that it will go the other direction.
---As traders, I would estimate that better than 90% of us believe that the market is neither random, nor efficient. If we didn't believe the market had directional bias, it would be very difficult to believe that we had found a "system", or an "edge" on the market.
---My desire to examine this contrast between traders and economists was piqued recently by a thread on BigMikeTrading started by Big Mike himself here:
I want to test something. It's simple. Place some random lines on your chart prior to the day opening, and see if at the end of the day you feel like those lines were important (try to imagine they weren't random, but some expensive or complicated …
where the forum examines putting random lines on a chart, much like fibonacci retracement lines, and then observe how the price seems to "react" to the lines. Randomness was assured by using a random number generator prior to the open, and using an assumed range. The results were remarkable, with many randomly generated lines acting as "Support" or "Resistance"; ideas that many of us find critical to our trading strategies.
---Among the lively discussion on that thread, there were many interesting optical illusions, but the one most interesting to this journal was posted by Fat Tails here:
I have just posted the illusions, because for me this is the essential message of the random lines:
Our brain interprets random events as having a cause.
Examples:
Random Lines: What ever line you draw, with hindsight it may look, as if it had an …
.
---In the post, there are multiple charts, some of which are randomly generated, and some of which are actual markets. I recall having seen this demonstration earlier in my trading life, but they held my interest longer this time. (Perhaps tempered by more humility than I had when I first aspired to dominate the markets with my superior intellect.) Some may see minor differences that indicate which is which to a highly experienced trader, but for a majority, I would guess that it would be difficult to sort them with any accuracy.
---Finally, I was reading a thread on elitetrader forum, which I find to be informative, but frequently devolves into name-calling and harassment. In this post, a frequent contributer is insulting another trader's system because it aims for many small wins, but leaves the stops wide. Mike, If I can't post the quote here, let me know and I will remove it; but here it is
---This concept of narrow stops with wide stops seems to be wide spread in the trading culture, and I decided to use this month to test it.
---If, indeed, the market is Random, a method that uses narrow profit targets should provide better results than one with narrow stops, for the simple reason that it is more likely to hit the narrower target and reverse than to continue all the way to a wide target.
And thus, my system for this months journal...
(I'll try not to make too many posts as verbose as this one! (maybe just the next one though...))
Andrew Lo is at MIT and a lot of his stuff is available to read online.
FWIW some people do both... the Random Walk/MPT portfolio that they feed regularly and rebalance once a year, and the "active" portfolio where anything goes. @Private Banker mentioned the "core-satellite" strategy a while back, which I know at least one of the big european PBs here liked to use for its HNW clients. The client could keep a finger in the market with the "satellite" portion, and (generally) sleep well at night knowing the "core" was safe.
---The idea, then, is to provide a means for testing different Profit Targets and Stoploss orders in a totally random trade.
---Why does it need to be a random trade?
Because if I don't randomize the trade, I run the risk of introducing directional bias from a system (Fibonacci, Trendlines, Oscilators, Volume studies, or Candlestick patterns to name a few) which may cause the results to be skewed one way or the other. The goal is to find the effectiveness of stops without an edge. If they work without and edge, then (in theory) they should work even better with an edge.
The System
Randomly assign two times to trade during the regular trading day. 1 from 8:30-9:15 CST and 1 from 2:00-2:45 CST.
---For this, I will use the website: RANDOM.ORG - Clock Time Generator to generate one time during morning peak trading times for high volatility, and another time at the afternoon peak, but leaving enough time to exit the position before trading ends. (see attachment 1) The times were chosen due to volatility, which can be seen in attachment 2, the Average change in the ES contract for each 15 minute period over the last 6 months.
Randomly determine which direction to enter the trade before market open to avoid bias.
---As stated in the opening post, random entry is a requirement. This will be determined by coin flip or internet RNG.
Place trades in the ES and YM in the same direction concurrently, with opposing Stoploss/Profit Target goals. Varying which trade is placed in which market each day, and each trade.
---The ES was chosen because it is notorious for trapping small stop traders in their positions, hitting their stops, and then reversing. While it trends, most simple trend following systems seem to be crushed be the Emini S&P.
---Because it would be difficult to track which stop was hit and which profit target was hit in NT7, I have decided to enter the two trades on different markets that trade essentially the same. Over the course of the month, their should be approximately 80 discreet R/T trades, 40 of each Stop, with 20 in each instrument, so the aggregate difference should show the method, not the market. (image three shows the current differential between the S&P and the Dow, and the fact that the relationship remains fairly constant over short time periods)
Place stops and targets based on the expected range for a 15-30 minute trade
Profit-Target to Stop Ratio of 2.5:1 trade 1, and 1:2.5 for trade 2
---Attachment 2 shows that the ES has an average range of 3.5+-1.0 points during the open, and 2.5+-1.0 during the close. Therefore, to keep the trades to short time frame, I will place the initial targets at 10:4 ticks
---The YM trades at a rough multiple of 9.2 to the ES, so the target:stop ratio will be 23:9 for the YM
I have considered the idea of trailing stops, and if I haven't driven myself crazy by the end of March, I will attempt to look into that that as well.
What inspired me to do THIS as my journal?
I have always been lax with my stops because I harbor a closet suspicion that they cost me more than I save.
I am sceptical by nature, and although I trade a reversion to the mean at some times, and a trend at others, I wonder if my own analysis isn't more akin to voodoo then science.
Mike has made me aware of a sick, hidden desire to have a tablet PC, even though they can't handle the vast majority of what I do on a computer. At least I would look good at Barnes & Noble with my Cafe!
If anyone actually reads this journal, I would be happy to hear any comments on my idea's effectiveness, or anything at all!
Looking forward to read you! Like the scientific approach!
Good luck in trading,
ivan
The markets are non-linear, dynamic systems. Chaos analysis has determined that market prices are highly random with a trend component. The amount of the trend component varies from market to market and from time frame to time frame.
Well, in a freak coincidence, the random time generator picked a perfect market reversal. If this happens regularly, I might have to just use random number generation as a tenet of my trading!
Another coincidence... My wife was away this morning, so I had to take my son to school just as the Market was opening. So, I was running out the door as I was posting and programing the first entries, and I wasn't there to watch the position open. Of course, I wouldn't be trading a strategy that I wrote in 2 minutes on a real account, which is good since I made several errors.
Error 1 -- I made the time based strategy >= and not ==. (attachment 1) Therefore, the strategy continued to enter the market every time it exited with a stop. This morning. This error would have made me lots of money! (attachments 3 & 4 for traded contracts) It also makes a solid case for Trailing stops, and based on those entries during a trend, it certainly makes the wider stop with a narrower profit target look better, as it didn't stop out once, and the wide profit target may have actually shown a net loss even in a market that favored it due to constant stopping out. Of course, those trades don't count, since that wasn't the trading plan.
Error 2 -- In running out the door, I forgot to enable the strategies in the NT7 control center. (attachment 2) However, since the fills are not in question here, even with slippage, I will accept them, only not with the extra profit from waiting until the close of the bar as shown in the screen shots, but at the profit targets.
@MXASJ, Thanks for the book idea. I will take a look at it next time I'm at the bookstore, but the reviews on Amazon looked interesting. I'm going to take a look at the Private Banker's "Core-satelite" strategy thread right now.
@kingich, Thanks for stopping by and wishing me luck. Hopefully, I'll learn something!
---Well, the trades executed correctly this afternoon in NT7, so the errors from the morning were resolved. As planned, the two instruments were swapped for the afternoon trade, and as hoped, the trades would have had the same result had they both been executed in the ES. I anticipate that there may be a few instances where the results will NOT be exactly the same, but I will try to note those as they occur. For the purposes of this test, it appears that they will work.
Itcame within .25 points of hitting its stop before turning and hitting its profit target. Obviously this was a risky trade against the trend, but since it was randomly generated and had fixed stop rules, it played out. As an aside, I had a hard time NOT feeling the stress of the trade, even on SIM, and I am fairly sure that I would have cut it short and made it a loser or at least break even in actual trading.
Score one for following Rules. Total to date: + $87.60 / ct
Wide Target/Narrow Stops : -$49.20 / ct
-9 pts YM * $5 = -$45 -$4.20 rtc
The trade was against the trend, although I suppose some might have hoped for a bounce off of prior support and jumped the gun, trying to maximize a reversal. This is not how I like to trade, but that's why it's random!
As such, the stop-loss behaved exactly as desired, and kept the position from being a big loser if it had gone further against us.
The price action went well past our stop before reversing and heading higher.
As of right now, 30 minutes to market close, it never reached our profit target and was heading lower. Total to date: +$72.60 /ct
---So, after day one, the narrow profit target has a small lead. We will see where tommorow leads us, but I did find a few interesting ideas while watching today.
I may rewrite the indicator to do all of the random generation from within the ninjascript. The possitive to this is that I don't have to get blisters from flipping coins and pressing the enter key on Random.org . The negative is that the setups would not be posted before the market opens. If anyone actually cares, I'm all ears. (litterally... even my mom made fun of 'em)
I noticed that the mistake in writing the indicator, where it continued to re-enter positions, led to another test all together. Most of us don't only make 2 trades a day, and we often re-enter the market after a position if we see another confirmation. Perhaps, as the month goes on, I will re-engineer the test to make multiple entries based on a random time generator cycled after the exit from the first trade, but still staying within the high range/volatility time periods. Ideas?
Watching a simulated trade that I am going to journal elicits almost as much emotion for me as an actual cash trade. I found myself routing for my sim trade at one point, even though it was in opposition to my actual position! For this, I may need a psychologist!
I hope everyone had a profitable, or at least educational, trading day. I'll post tomorows setups later tonight.
Excellent job on the journal. I'm also happy I piqued your interest with the random line thread, and that you "got it" and understood what I was trying to demonstrate. I hope it was helpful, and I look forward to your journal!