Sunday, September 14, 2014

How I Use Tradervue with BOWS Bootcamp

Last month, I enrolled into the Bulls on Wall Street (BOWS) 60-day Bootcamp training program for daytrading stocks, and it has surprisingly exceeded my expectations.  After years of studying trading and joining various chatrooms, I initially thought, what else could I really learn or gain?  

Before joining, I asked several prior students and they all said the same -- the Bootcamp was well worth the time and cost.  And yes, I was still a bit skeptical, I thought I was different.

I was wrong.  

I've realized that you never stop learning something new, even if you think you already know a lot. A surprise value-add has been the highly focused / low noise content within the private Bootcamp Google+ community from all the current and prior students, where I continue to learn and also try to help pay it forward.

I'll write more about the Bootcamp and why I joined in a future post.  

One interesting and reassuring topic within the Bootcamp has been the use of Tradervue for journaling trades.  I've written much about Tradervue in the past here and it has been great to dust off some of my old Tradervue tags from back when I traded stocks more actively.  

Posted below is a slightly modified crosspost I made to the private BOWS community regarding how I'm using Tradervue now to create and refine my trading plan.

* * * * *

"Better decisions through data" was the motto for a company I used to work for many years ago.  It's something that I still strongly believe.  And analyzing data has been a big part of helping me to make better trading decisions.

In this post, I'll explain the Tradervue analysis of my SIM (simulated) trades to determine whether the setups I outlined in my initial trading goals are still applicable.  

Based on the performance of the setups, I will then create and refine my trading plan.  This is working somewhat "backwards", but given my trading background, this process works for me.

Trading in the Clique Fund (BOWS hedge fund) is my goal, so the trading conditions (trading share size, daily loss limits, etc.) I used are similar to initial conditions when I go live.  

However, I still chose to measure my performance on the R-multiple scale within Tradervue, which is simply a ratio of how much you made/lost compared to what you initially risked. 

a) Risk $10 and make $20, you made +2.0R
b) Risk $10 and lose $5, you made -0.5R

That way, I focus on reward/risk for every trade, NOT absolute $'s.  It's easy to brainwash yourself and become overconfident with big fat $ returns under SIM, only to get smacked hard with reality when you go live.  

It's also harder to "fool yourself" with the R-multiple returns.  For example, you can make up a big losing day with one last huge unrealistic 50,000 share winning trade, whereas the R-multiple returns will still likely show you with a losing day.  

I've fooled myself too many times in the past, so I've learned some lessons.  Use Tradervue or any other journaling tool/spreadsheet to look at the cold hard facts.  

It was a decision filled with pros/cons, but I decided to reuse many of the tags from my prior experiences in stocks.  Therefore, some of the tags you see in this analysis may not be a part of the BOWS vernacular.

Once the tags are entered for each trade, it's very simple to filter the results to only view or report the performance for a specific subset of accounts.

To only show or report on trades that were long bullflags that were in a ascending triangle, I would type in the tag field:

       long AND bullflag AND asctri


*  Results are based on SIM trades (beware of the *fantasy fills*!), gross (no commissions), 25-100 shares per trade (with few rare 200 share trades), $150/gross daily max loss.  My goal is to get used to initial Clique fund conditions.

*  R-multiple return was +59, based on 400 trades from 8/19/2014 to 9/12/2014.  "Probability of random chance" = 4%, so results are statistically valid.

*  If a consistent 1% of portfolio was risked per trade (R=1%), that's an almost 60% return.  And if 2% was risked per trade, it's almost 120% return. But this is NOT a realistic assumption -- it's hard to daytrade with a consistent R value.  Reality for me will likely be more like 0.5%.

*  I overtraded, surprise!  And only trading 100 shares on lower priced stocks, even if in super-juiced-momo-mode, is very challenging to make decent net dollar returns consistently.  Commissions and errors also start to compound dramatically with overtrading.

*  Winning accuracy was about 50%, which is not great given the lower profit factor value.  I was not as selective as I could have been.  In the back of my mind, I was thinking, "I'm collecting data, so I'll just do this to see what happens."  Not a good habit to make.


Under my initial goals, I thought focusing on flags, ORB, and 123 tops/bottoms would generate the best results.  

WRONG!  Well, kinda...  Here are some key findings.

*  The long side performance dominated over the short side.  Perhaps that's just the current condition of the markets, as well as bias of stocks.  Or maybe the chatroom's bias.

*  Therefore, the bull flag and 123bottom setups also significantly outperformed the short side of those respective setups.

*  I always thought the bearish "h setup" was a good setup for me, but based on the past month, that's WRONG.  Although that setup might do better if overall markets go into correction mode.

*  The ORB setups were profitable overall, but didn't generate a decent average R per trade.  In addition, some of those ORB setups should have gotten runners that run the entire day. Likely I'm doing it wrong.

*  The triple tap breakouts as well as the sideways consolidation type flags were surprise solid performers.  However, need bigger sample for the triple tap breakouts.

*  I thought I'd be a pretty good trader of 1min charts by now, but the data clearly says STAY AWAY.  I need to focus primary on the higher timeframes, and rarely use 1min except when absolutely necessary.

*  Revenge and rogue type trades were a small percentage (< 4%) of overall trades.  But still, total R-multiple profits could have been about 20% higher without them, these trades are a huge negative impact!

*  There is an opportunity to do an even deeper dive to optimize what combination of tags do best, as well as time of day type analysis, and more. This will be something to explore in the near future. 

*  I will need to focus on creating tags to help determine how best to optimize exits.  Exits are arguably more difficult than entries.  

*  Additional reports from the upgraded version of Tradervue have been attached.  These are just a very small number of interesting reports that could be used to further refine your trading, I've only scratched the surface.  Most of the reports should be self explanatory, but please feel free to leave questions in the comments.
A trend report of average P&L
Determine optimal stop amount
Trend of average heat taken on trade (MAE)
Evidence of trade mis-management -- letting winners turn to losers
Report based on tags

Now that I have a clearer idea of what works and what doesn't, the findings from above will be used to modify and tighten up my initial trading plan.  

Focus on your strengths is a common mantra. Therefore: 

I will most likely focus my trading plan on the setups with the magenta values under the PF (profit factor) column of the INITIAL PERFORMANCE BREAKDOWN report.  
Why should I use a lot of energy to try and become a profitable 1 minute chart trader when I'm already doing pretty well with other setups?  

Markets ebb and flow -- what works now may not work in the future and vice versa.  So all these setups will continue to be monitored, and as the market changes, adjustments will most likely be made in the future.

Friday, September 5, 2014

Response from Spread The Trend

As was clear in my last post, as well as some of the comments under the Spread The Trend review (originally posted in October 2013), there have been some rough patches recently.  This happens to all traders, funds, investors, etc.  Participate in the markets long enough, and no one is immune to experiencing challenging times at some time or another.

Spread The Trend replied under the comments of the original review.  I thought it would be good to promote it to a blog post, since there are some clarifications, explanations, and admissions that will be of value to some readers.

Here is the comment in its entirety:
We are responding to the comments made by Marco Starr whose name is unfamiliar to us. When a trade is rolled, you close one position and then simultaneously open another position. The losses being referred to are paper losses. A true realized loss is not incurred unless someone chose to simply close all positions and not roll their trades. We are currently down for the year with implied volatility being anemic but our final tally is far from solidified. We have also been trading in and out of several positions to help reduce the cost of our rolls and being as conscientious as possible to our subscribers.  
In regards to “wiping out several years of gains” this seems to be an over reaching statement. You only need to view our track record page to see that we are well ahead of the game if you consider the longer and even medium term. Now had someone only been with our service for less than a year, then yes it is possible that their account could've temporarily dropped below the value they started with.  
All along it was explained to subscribers what our plan of action was and why our decisions combined with market forces lead to these outcomes. In all fairness, perfection in trading is difficult and we stand by our longer-term track record. Of course, we do our best to try and avoid getting into stressful situations altogether. We also believe we could have done a better job as well and have taken “key learning’s” to heart from this experience.
The majority of auto trade subscribers, or over ¾, did fill on the roll in question. As for those folks trading their own accounts the results seemed mixed as some did fill and some did not. It’s most common for everyone to fill or not fill as a whole but on rare occasions sometimes not everyone will get filled all in the same day. This can possibly be due to the timing of the order being sent or how quickly the broker is able to submit the order, or it could even have something to do with the liquidity in the market at the time the order is processed. 
When we see that the mass majority is filling that tells us that our order is priced fairly and correctly. If a broker did not fill, then we ask the question why? There are tools available where one can go back and see the amount of volume traded on these strikes, which shows there were many fills. Could Spread The Trend have sent the order earlier? In hindsight, probably yes. However, the process that has been successful for many years dictated that waiting was appropriate. We desired a different outcome as well and these trades impact our own personal accounts.  
All trades are posted on our website in the Member's section under current open positions with all of the details in their entirety. We do not post the trades to our closed positions page until the entire trade has been closed out and as of Mr. Starr’s posting and this posting; the entire trade has not yet been closed out. 
We offer a one month $1 trial so folks can get comfortable with the strategy and always ask that individuals add capital gradually. One may review both open and closed trades and decide if the newsletter is a good choice for them. Spread The Trend realizes that this period was stressful but we also have many winning weeks, months, and years on record. However, that concept that we were a “deer in headlights”, “virtually sunk”, or “listing” is disingenuous to our long term subscribers who are somewhat disappointed but are still extremely profitable and content with our strategy.