Stock market realization

After 2 weeks of watching technical stock market analysis videos, Brian Shannon over at Alpha Trends has broken through my thick skull and opened my eyes to probably a much more proper way to view my trading. As more of my posts reveal (here and on my other blogs such as Financial Jetstream), I’ve been trading for a couple years now… with, I suppose more or less the idea that I should really (REALLY) try to do extensive research in to a stock before getting in. And then I would get rather frustrated with my lack of skills when the stock I’d picked went against me. I know I’m new at this, with a whole lot more to learn, but some times it just seems like no matter how much work/study I do, it will not be enough. A lot of that also comes from me watching other traders who ARE a lot more experienced continually make wrong predictions. And all these traders/analysts are saying different things about the market or even the same stock. It’s enough to make me want to pull out my hair!

However, today, it clicked… as Brian gave some insight to what he thought two stocks might do in the near future. He tossed out the idea for a stop, and also a target price, and the key words being “Here we have 17 cents of risk, for a potential gain of 90 cents”. On the next stock, the same words “Here we have 22 cents of risk, for a potential gain of $1.40″

Whoa! There it was! The idea that we don’t really know for sure which way the market will go… ever.. no matter how good you are, and perhaps picking stocks could be similar to flipping a coin (well, hopefully slightly better odds), but if the usual loss is 20% while the gains are 80% or more, the odds are WELL in your favor of coming out far ahead of the game (even on that 50/50 coin toss).

To prove this point I wrote a short little Perl script:

perl -e '@a=(10000,10000,10000,10000,10000); $t=1; while ($t<500){ foreach $a (@a){ rand()<.5 and $a*=.8 or $a*=1.3; $a<50 and $x=1} printf "%3d = %12.1f %12.1f %12.1f %12.1f %12.1f\n",$t,@a; $t++; $x==1 and exit }'

which simulates 5 accounts starting at $10,000 each, and then makes 500 trades with each of them. There is a 50% chance to lose 20% of the accounts value and a 50% chance to gain 30% of the accounts value during any trade. I ran this script 40 times to simulate 200 accounts doing this trading, and found that only 5 times in 200 did the account value drop below $50. To me, those are pretty good odds of success on a supposed 50% chance of gain/loss per trade. What makes this work is that when you do win, you win more than you’d lose… and THIS simulation is only with a win of 30%.

Just for fun I created another script:

perl -e '$loss=.0116;$win=.0496; for ($c=0;$c<100;$c++){$a[$c]=1000}; $t=1; while ($t++<1000){ foreach $a (@a){ $a==0 and next; rand()<.5 and $a*=(1-$loss) or $a*=(1+$win); $a<50 and $a=0} }; $bust=grep($_<100,@a); $loss=grep($_<=1000 && $_>100,@a); $fifty=grep($_<=1500 && $_>1000,@a); $double=grep($_<=2000 && $_>1500,@a); $more=grep($_>2000,@a); print "BUST: $bust\n","LOSS: $loss\n","FIFTY: $fifty\n","DOUBLE: $double\n","MORE: $more\n";'

Which sets up 100 accounts with $1000 each, then makes 1,000 trades in each of them. Each trade has a 50% chance of failure or success where it will lose a smaller amount than it will gain on a win.

Test #1: (the $23.15 stock with $0.27 risk (-1.16%) or $1.15 gain (+4.96%) after 1000 trades in 100 accounts are:
BUST: 0 (left with under $100)
LOSS: 0 (left with from $100 - $1000)
FIFTY: 0 (left with from $1000 - $1500)
DOUBLE: 0 (left with from $1500 - $2000)
MORE: 100 (left with more than $2000 - double the start)

Test #2 - now with only a 3% gain … after 1000 trades in 100 accounts are:
BUST: 0 (left with under $100)
LOSS: 0 (left with from $100 - $1000)
FIFTY: 0 (left with from $1000 - $1500)
DOUBLE: 0 (left with from $1500 - $2000)
MORE: 100 (left with more than $2000 - double the start)

Test #3 - now with only a 2% gain … after 1000 trades in 100 accounts are:
BUST: 0 (left with under $100)
LOSS: 0 (left with from $100 - $1000)
FIFTY: 0 (left with from $1000 - $1500)
DOUBLE: 0 (left with from $1500 - $2000)
MORE: 100 (left with more than $2000 - double the start)

Test #4 - now with only a 1.5% gain … after 1000 trades in 100 accounts are:
BUST: 0 (left with under $100)
LOSS: 0 (left with from $100 - $1000)
FIFTY: 1 (left with from $1000 - $1500)
DOUBLE: 2 (left with from $1500 - $2000)
MORE: 97 (left with more than $2000 - double the start)

Test #5 - now with only a 1.3% gain … after 1000 trades in 100 accounts are:
BUST: 0 (left with under $100)
LOSS: 6 (left with from $100 - $1000)
FIFTY: 22 (left with from $1000 - $1500)
DOUBLE: 28 (left with from $1500 - $2000)
MORE: 44 (left with more than $2000 - double the start)
…. which means you STILL only have a 6% chance of losing any of your principle and almost a 50% chance of more than doubling your money.

After these tests I can obviously see a better picture of why the risk-to-reward numbers are SO vital, and it’s not something I was taking in to account on my trades in the past.