BWolinsky Trading
- Thread starter bwolinsky
- Start date
- Status
As always, and I wouldn't dispute any of that. You're correct in the "what have you done for me lately", but I still point out that waiting for more opportune times is always the best solution, and why I believe in trading QID and QLD with Pairs Trading QID QLD 2.0.
The Cash Cow system is another story. It has severly lagged the market, and I know it can go up 50% in short periods of time, but I'm wondering if it will happen this year or not.
I'm pretty excited about this week. The move up after hours tonight I believe is reflective of the anticipation of better than expected retail sales for the holiday weekend. Virtually any increase in holiday sales would beat any consensus estimate on the Street, currently, so banking on a quick trade in the release of that number will be a pretty good bet if it comes out better than expected.
http://www.bloomberg.com/markets/ecalendar/index.html
Specifically pay attention to the Chain Store Sales number, as it would be a leading indicator for future sales reports, especially from retailers. I think you can pretty much extrapolate from that number what other retailers did, and, as I said, any increase would cause a market rally in my opinion. Look for any news that talks about the sales over the weekend, and I think that'll be mostly what drives us this week. I think we'll get some reports of good and bad areas, but, as a whole, once we see the chain store sales number, I think it will be much better than expected.
I'm planning on reviewing the old June 2009 Level II CFA Exam material I have, and studying till I register for the June 2011 exam in a couple years. My failure on the exam I don't believe was a lack of preparation. I believe my failure specifically was in only preparing based on Schweser Kaplan notes and watching the classes. Practically none of my time was spent reading the actual CFA Institute Curriculum Vitae. My loss, I guess, but I never read any of the CFA books for Level I. None, I actually returned them to the CFA Institute in the year I passed, and the year after that, they made it mandatory for CFA candidates to purchase $300 of books in addition to the $800 registration for the exam with no returns allowed.
SO... I will come back to it. There were certain questions I recall on the exam that seemed to be taken off of very specific examples only written in the actual CFA Institute textbooks. Again, Level I is basically plug and chug. You take two 120 question 3 hour exams at level I, and it tests speed predominantly. At Level II, you only have to answer 60 questions in each 3 hour session. The format at Level II is another reason for my failure. The Case Studies or "Vignettes" required all previous answers to be correct before having a chance to answer later questions. If you got stopped at question 3, pretty much you couldn't answer the next 3 questions, and basically guessed answers at that point. I found that the first 3 questions were always easy, and every 4-6 question number was tied to the inital 3 questions in some way, at much higher difficulty levels. I aced Corporate Finance, Alternative Investments, and Derivatives with greater than 70%, but I knew they were all correct. The rest were below 70%, but really it's not a very good measure of reality. If there were only 6 questions, and you only got 4 right, automatically you're below 70%, and I was always banking on having at least 4 out of 6 right. If I had that I think I would have passed. Scoring a 3 out of 6 puts you at 50% and below. There was a lot more below 70% than above, but being 1 in 533 to be a member of the CFA Institute should tell you something. There are only 53k current CFA members, and only 45000 CFA Charterholders, approximately.
As I said in another thread, I have yet to be denied an interview for any investment advisory, portfolio management, or research analyst position, because I have Level I of the CFA Exam under my belt. Each time I asked why I wasn't hired, I received the response of a 10 year veteran of Wall Street being chosen over my credentials. Well, some of these positions were just exciting to have actually been considered for. I went out and bought some suits for a couple of them, deducting costs from my business in the process.
I'm looking forward to the rest of the year. I can't wait to see 2010 roll around. I think the market will breathe a sigh of relief to be out of this year. I don't know of any financial catastrophe on the horizon. I am convinced that I will not have to deal with 56% drawdowns in the market for the rest of my life. Maybe some 40%'ers, but never 56%. It was "too stupid for words", and the whole time you could tell by the median price to book ratio of the market being below 0.7, when it's usually greater than 1.4.
Onward and upward.
http://www.bloomberg.com/markets/ecalendar/index.html
Specifically pay attention to the Chain Store Sales number, as it would be a leading indicator for future sales reports, especially from retailers. I think you can pretty much extrapolate from that number what other retailers did, and, as I said, any increase would cause a market rally in my opinion. Look for any news that talks about the sales over the weekend, and I think that'll be mostly what drives us this week. I think we'll get some reports of good and bad areas, but, as a whole, once we see the chain store sales number, I think it will be much better than expected.
I'm planning on reviewing the old June 2009 Level II CFA Exam material I have, and studying till I register for the June 2011 exam in a couple years. My failure on the exam I don't believe was a lack of preparation. I believe my failure specifically was in only preparing based on Schweser Kaplan notes and watching the classes. Practically none of my time was spent reading the actual CFA Institute Curriculum Vitae. My loss, I guess, but I never read any of the CFA books for Level I. None, I actually returned them to the CFA Institute in the year I passed, and the year after that, they made it mandatory for CFA candidates to purchase $300 of books in addition to the $800 registration for the exam with no returns allowed.
SO... I will come back to it. There were certain questions I recall on the exam that seemed to be taken off of very specific examples only written in the actual CFA Institute textbooks. Again, Level I is basically plug and chug. You take two 120 question 3 hour exams at level I, and it tests speed predominantly. At Level II, you only have to answer 60 questions in each 3 hour session. The format at Level II is another reason for my failure. The Case Studies or "Vignettes" required all previous answers to be correct before having a chance to answer later questions. If you got stopped at question 3, pretty much you couldn't answer the next 3 questions, and basically guessed answers at that point. I found that the first 3 questions were always easy, and every 4-6 question number was tied to the inital 3 questions in some way, at much higher difficulty levels. I aced Corporate Finance, Alternative Investments, and Derivatives with greater than 70%, but I knew they were all correct. The rest were below 70%, but really it's not a very good measure of reality. If there were only 6 questions, and you only got 4 right, automatically you're below 70%, and I was always banking on having at least 4 out of 6 right. If I had that I think I would have passed. Scoring a 3 out of 6 puts you at 50% and below. There was a lot more below 70% than above, but being 1 in 533 to be a member of the CFA Institute should tell you something. There are only 53k current CFA members, and only 45000 CFA Charterholders, approximately.
As I said in another thread, I have yet to be denied an interview for any investment advisory, portfolio management, or research analyst position, because I have Level I of the CFA Exam under my belt. Each time I asked why I wasn't hired, I received the response of a 10 year veteran of Wall Street being chosen over my credentials. Well, some of these positions were just exciting to have actually been considered for. I went out and bought some suits for a couple of them, deducting costs from my business in the process.
I'm looking forward to the rest of the year. I can't wait to see 2010 roll around. I think the market will breathe a sigh of relief to be out of this year. I don't know of any financial catastrophe on the horizon. I am convinced that I will not have to deal with 56% drawdowns in the market for the rest of my life. Maybe some 40%'ers, but never 56%. It was "too stupid for words", and the whole time you could tell by the median price to book ratio of the market being below 0.7, when it's usually greater than 1.4.
Onward and upward.
QLD Projection: 60.0180454463749 QLD Close: 54.8600006103516
QLD Projection: 53.769491854581 QLD Close: 54.8600006103516
QLD Projection: 54.6088389805385 QLD Close: 54.8600006103516
RSQ1: 0.787402391433716
RSQ2: 0.794829308986664
12/1/2009:QLD Predict: 59.2119441838055 QLD Close: 54.8600006103516
<b>Norm is True on 12/1/2009</b>
68.1125439624854% of the time we, have been fairly valued
The good news, which I have highlighted in bold is that we are now "normalized." This means that the primary overbought and oversold thresholds have been activated, and any breach of these levels will trigger a trade in QID or QLD.
I'm hoping the market comes in a bit before the Chain Store Sales number is released. I think we'll get a big boost up from that number if it shows any growth in sales Year on Year. Tomorrow is Wednesday, and the sales number's on Thursday. Hope for a drop, breach the oversold thresholds, and it looks like we are likely to enter on Thursday morning, because we are very close to the actual threshold, but had to cross over it to have normalized volatility as I call it. Being just slightly above the oversold threshold puts us on the lower end of a fair value range, but not quite past the level we need to be at to take a position.
So, the good news is that we'll take any regular trade from here on out until we have a loss, which is my trigger for both secondary threshold levels and wash sale contgency algorithms mainly there for tax considerations and as a check against standing in front of a trending market rather than continuously taking a countertrend trade.--
Good Trading,
Good Luck,
Beau Wolinsky
QLD Projection: 53.769491854581 QLD Close: 54.8600006103516
QLD Projection: 54.6088389805385 QLD Close: 54.8600006103516
RSQ1: 0.787402391433716
RSQ2: 0.794829308986664
12/1/2009:QLD Predict: 59.2119441838055 QLD Close: 54.8600006103516
<b>Norm is True on 12/1/2009</b>
68.1125439624854% of the time we, have been fairly valued
The good news, which I have highlighted in bold is that we are now "normalized." This means that the primary overbought and oversold thresholds have been activated, and any breach of these levels will trigger a trade in QID or QLD.
I'm hoping the market comes in a bit before the Chain Store Sales number is released. I think we'll get a big boost up from that number if it shows any growth in sales Year on Year. Tomorrow is Wednesday, and the sales number's on Thursday. Hope for a drop, breach the oversold thresholds, and it looks like we are likely to enter on Thursday morning, because we are very close to the actual threshold, but had to cross over it to have normalized volatility as I call it. Being just slightly above the oversold threshold puts us on the lower end of a fair value range, but not quite past the level we need to be at to take a position.
So, the good news is that we'll take any regular trade from here on out until we have a loss, which is my trigger for both secondary threshold levels and wash sale contgency algorithms mainly there for tax considerations and as a check against standing in front of a trending market rather than continuously taking a countertrend trade.--
Good Trading,
Good Luck,
Beau Wolinsky
As I was working out tonight to prepare for my debut back into Ice Hockey, something dawned on me that I should have said some time ago.
The background story is that I heard this from my wife. As one of the greatest Hallmark Designers, my wife has some professional priveleges that she uses quite a bit. It just so happened that she attended an Oprah University event appropriately called O U. The event hosted in downtown KC had several key note speakers, including Suze Orman, Dr. Ozz, and Stacey London.
One of these keynote speakers mentioned a childhood story about a friend of theirs who at age 13 called himself, "The Best." Naturally, this afore mentioned keynote speaker promptly replied, "No, you're not." The name of this individual was from South Side Chicage where the speaker grew up, and I will save the actual names of these people for the weekend. I'm sure when you know who it is, the story will make perfect sense to you, but not without some debate first.
It's no secret my Covestor profile states I'm one of the best trading strategy developers in the world, and obviously I believe that's true. Why I think that is true is based on years of strategy trading and fine tuning of trading strategies. No one has published more profitable trading systems than I have, and I also don't believe anyone currently possesses more robust systems than I have. Whether those systems currently reflect what I believe their intrinsic performance characteristics are is of no consequence, because eventually, they will be profitable. Over any arbitrary short period of time, they may not appear as robust as first blush, but that doesn't mean they never will.
Getting back to the topic at hand was that the speaker spoke of this individual believing he was "The Best" at a very young age, 13, in fact. Naturally the speaker was skeptical, but the logic of this individual, who shall remain nameless until this weekend, was that you have to first <i>believe you're the best, in order to <b>become</b> the best.</i> Exactly the same logic I apply to my daily professional activities. It's up to you to decide for yourself based on years of internet based posts, and after several published trading systems still verified by third parties to be extremely profitable as well as a walk forward history to examine as I progress. Currently you see me near the beginning of what I might call the "third party verified, real financial results." Since you cannot see my actual trades prior to joining covestor, unfortunately, you'll just have to take my word for it that I was trading during that time. It should be obvious from a cursory comparison of the original Pairs Trading QID QLD Scalper and my covestor trades that they were nearly identical. Thinking I wasn't trading is stupid, because I would not still have Wealth Lab Pro at my disposal if I wasn't, and would not have been able to continue my strategy development.
The point I'm making is that my bravado is not without merit. What you see currently is only a tiny speck of what I expect myself to accomplish in the next several years.
<i>I believe I'm the best because I feel I am becoming the best. If you don't think so, try putting yourself out there for public scrutiny sometime, and you'll find it takes a lot more confidence and ability to ignore naysayers than it appears.</i>
So, the game I'm going to play is guess who?
Round 1, figurative $2,000 question)
Which speaker was I referring to?
Grand Prize, figurative $64,000 question)
Which individual was this speaker talking about?
I'm sure people will get round 1 if they just think about the story a bit, and possibly have heard this speaker refer to it on television already....but if you have, kudos if you get <b>both</b> correct, because it means you were paying attention.
Good Trading,
Good Luck,
Beau Wolinsky
The background story is that I heard this from my wife. As one of the greatest Hallmark Designers, my wife has some professional priveleges that she uses quite a bit. It just so happened that she attended an Oprah University event appropriately called O U. The event hosted in downtown KC had several key note speakers, including Suze Orman, Dr. Ozz, and Stacey London.
One of these keynote speakers mentioned a childhood story about a friend of theirs who at age 13 called himself, "The Best." Naturally, this afore mentioned keynote speaker promptly replied, "No, you're not." The name of this individual was from South Side Chicage where the speaker grew up, and I will save the actual names of these people for the weekend. I'm sure when you know who it is, the story will make perfect sense to you, but not without some debate first.
It's no secret my Covestor profile states I'm one of the best trading strategy developers in the world, and obviously I believe that's true. Why I think that is true is based on years of strategy trading and fine tuning of trading strategies. No one has published more profitable trading systems than I have, and I also don't believe anyone currently possesses more robust systems than I have. Whether those systems currently reflect what I believe their intrinsic performance characteristics are is of no consequence, because eventually, they will be profitable. Over any arbitrary short period of time, they may not appear as robust as first blush, but that doesn't mean they never will.
Getting back to the topic at hand was that the speaker spoke of this individual believing he was "The Best" at a very young age, 13, in fact. Naturally the speaker was skeptical, but the logic of this individual, who shall remain nameless until this weekend, was that you have to first <i>believe you're the best, in order to <b>become</b> the best.</i> Exactly the same logic I apply to my daily professional activities. It's up to you to decide for yourself based on years of internet based posts, and after several published trading systems still verified by third parties to be extremely profitable as well as a walk forward history to examine as I progress. Currently you see me near the beginning of what I might call the "third party verified, real financial results." Since you cannot see my actual trades prior to joining covestor, unfortunately, you'll just have to take my word for it that I was trading during that time. It should be obvious from a cursory comparison of the original Pairs Trading QID QLD Scalper and my covestor trades that they were nearly identical. Thinking I wasn't trading is stupid, because I would not still have Wealth Lab Pro at my disposal if I wasn't, and would not have been able to continue my strategy development.
The point I'm making is that my bravado is not without merit. What you see currently is only a tiny speck of what I expect myself to accomplish in the next several years.
<i>I believe I'm the best because I feel I am becoming the best. If you don't think so, try putting yourself out there for public scrutiny sometime, and you'll find it takes a lot more confidence and ability to ignore naysayers than it appears.</i>
So, the game I'm going to play is guess who?
Round 1, figurative $2,000 question)
Which speaker was I referring to?
Grand Prize, figurative $64,000 question)
Which individual was this speaker talking about?
I'm sure people will get round 1 if they just think about the story a bit, and possibly have heard this speaker refer to it on television already....but if you have, kudos if you get <b>both</b> correct, because it means you were paying attention.
Good Trading,
Good Luck,
Beau Wolinsky
Just began my first reading in what was the June 2009 CFA Curriculum. The subject matter was just a review of Ethics, and there are three more readings to go in the first session.
To other matters.
I feel extremely confident in the results for Pairs Trading QID QLD 2.0.
I have been working towards applying the lessons learned from wash sales, autostops, and stop loss trades and have come up with a model for SDS and SSO. I also was working on FAZ and FAS, but it appears there is insufficient data to create a backtest with a statistically significant number of trades.
On the left is the SDS and SSO model, which I'm sure is obvious if you had followed my previous posts.
Long + Short Long + Short
Starting Capital $25,000.00 $25,000.00
Ending Capital $471,232.35 $1,081,660.08
Side by side, I can understand why someone would question the reason for the SDS and SSO lagging QID and QLD so much, but the annual returns show a different story, to me at least, having analyzed these reports for a long time.
Net Profit $446,232.35 $1,056,660.08
Net Profit % 1784.93% 4226.64%
Annualized Gain % 137.82% 203.89%
Exposure 74.79% 50.39%
Either one would still be a good system to use, till you see the DD.
Number of Trades 76 88
Avg Profit/Loss $5,871.48 $12,007.50
Avg Bars Held 4.91 2.8
Winning Trades 50 66
Winning % 65.79% 75.00%
Gross Profit $662,789.16 $1,727,862.95
Largest Winning Trades $74,588.57 $205,057.31
Avg Profit $13,255.78 $26,179.74
Avg Bars Held 5.08 2.86
Max Consecutive 7 10
A lot less trades with the SDS and SSO models. Hold times are also more than twice the QID and QLD version. Less of a win streak in the SDS and SSO models.
Losing Trades 26 22
Losing % 34.21% 25.00%
Gross Loss ($216,556.82) ($671,202.87)
Largest Losing Trade ($34,355.63) ($108,434.07)
Avg Loss ($8,329.11) ($30,509.22)
Avg Bars Held 4.58 2.59
Max Consecutive 4 2
As expected, a lot lower win percentage with SDS and SSO. Max consecutive is still inexcusable, but, by itself, not what I have identified as the cause of the Max DD.
Max Drawdown ($90,321.31) ($133,611.63)
Max Drawdown Date 8/11/2009 10/1/2009
Max Drawdown % -39.01% -20.34%
Max Drawdown % Date 10/8/2008 11/3/2008
So, here we are. Max DD for SDS and SSO is almost twice as much as the QID and QLD system. I attribute this to one trade, and one trade only. There is a trade in SDS and SSO that I can only call catastrophic. It is a losing trade that lost more than 17% in one day from 10/7/2008 to the Open on 10/8/2008.
Here is the tradeslist from that period. To show how volatility based overbought, oversold indicators adapt for changing market conditions, I have included the next two trades.
Position Symbol Shares Entry Date Entry Price Exit Date Exit Price % Change Net Profit Bars Held Profit/Bar Entry Signal Exit Signal Cum Profit MAE % MFE % ChartScript Delta Norm Series Actual Ratio Series
Long SSO 3,840 10/7/2008 41.53 10/8/2008 34.42 -17.13 -27,324.30 1 -27,324.30 Buy SSO Stop 29,766.54 -17.12 0.84
Long SSO 3,483 10/10/2008 27.09 10/13/2008 31.77 17.25 16,278.54 1 16,278.54 Buy SSO Profit Target 46,045.08 -11.04 19.23
Long SDS 1,538 10/21/2008 89.99 10/22/2008 100.47 11.63 16,102.24 1 16,102.24 Buy SDS Profit Target 62,147.32 -3.34 11.65
So, in defense of taking a loss like that, the next two trades were the number 1 and number 2 most profitable trades. I have attempted to find a way to "fit" this out, but mechanically, I have found no way and no other values to optimize the results any better, so I'm left to understand that to use SDS and SSO might warrant some possible attention to more than extreme volatility, and just plain irrational price action. I will need to ponder this for much longer, but the stats below give me pause as well.
APD 0.8301 0.7864
APAD 1.8742 1.7146
Oh the irony of having a system with 39% drawdown have a higher APD than a 20% drawdown. I can't make this stuff up. It's priceless, and a perfect example of what you'd go through to write 50 page long algorithms.
For those who don't know what APAD is, if APD is net profit over total drawdown, APAD is gross profit over total drawdown. The reason to look at APAD is to find whether your typical trade will overcome the total drawdown of the system, and I would say anything above 1 will.
Wealth-Lab Score 112.3872 322.2863
RAR 184.2791 404.595
MAR 3.5328 10.0224
Profit Factor 3.0606 2.5743
Recovery Factor 4.9405 7.9084
Before I discard the system, the returns below require some analysis. This WL score specifically might not accurately capture the underlying APR of the system as evidenced by the returns below.
Sharpe Ratio 1.8133 2.1658
My gold standard of high risk adjusted returns is always 2 on the Sharpe. Not that that will guarantee when you actually trade that you'll have a Sharpe of 2 the whole time, but that it has to be that high to <i>tolerate losses.</i>
Sortino Ratio 4.2993 5.4716
I have very rarely found any value to the Sortino ratio, because I truly believe backing out upside volatility masks upside drawdown. A true example of a system that has gamed the drawdown out of the equation through buy and hold is ETF timer. If you never sell a winner, you'll never have any drawdown, despite the position fluctuating 20-30k or 10-15% in a few days to weeks.
Ulcer Index 9.8026 6.1128
WL Error Term 29.287 8.755
WL Reward Ratio 4.7059 23.2885
Luck Coefficient 5.6269 7.8327
Pessimistic Rate of Return 2.1969 1.8607
Equity Drop Ratio 0.0508 0.0187
K-Ratio 0.3023 0.4032
Seykota Lake Ratio 0.0868 0.0449
Expectancy 0.6052 0.628
Expectancy Score 13.0956 15.9747
Max Losers Held 1 1
Max Winners Held 1 1
I don't find any value in the rest of these, but maybe someone out there will.
Now, to the part that I said I'd compare annual returns to. SDS and SSO is on the left, and QID and QLD is on the right
% Return % Max DD Period Starting % Return % Max DD
19.76 -6.22 7/13/2006 59.92 -10.19
-4.22 -28.55 1/3/2007 216.52 -12.33
370.79 -39.01 1/2/2008 237.8 -20.34
249.04 -24.5 1/2/2009 153.05 -16.23
So what does a professional see? I see a system that required a lot of data to be compiled before showing stellar returns. Every profitable system I've ever made has time periods of both stagnant and exponential growth. While I notice it took some time for the SDS and SSO version to come back, it has not benefited as much from the compounding effect the consistent returns generated by the QID and QLD system had, but that does not take away from the fact that the last two years the SDS and SSO version has significantly outperformed the QID and QLD version.
Now, I see confusion here. I know of quite a few funds that very long periods of time underperform, and in those few quarters of stellar growth essentially make the whole decade worth it for their investors.
I find trading to be a numbers game, with a lot more edge than card counting at Blackjack. Everyone will tell you to have a plan when you start, but they don't have any idea <b><i>how long it will take you to realize you need a better one.</i></b>
My approach to trading and investing is that I put myself from the perspective of passively buying an index, and benchmark myself to the S&P. By doing that, tolerating losses when the market goes down is much easier. In the periods where it goes back, I find myself always buying on the dip, which is another axiom prevalent in nearly every major trading site on the internet. It's certainly popular with professionals.
I liken the way I trade to a simple motto, and I know everyone has heard this before, but it's actually what I do with these pairs systems: <b>Buy when everyone's selling, and sell when everyone's buying.</b>
And that's all you have to do, no matter which timeframe your in. You'll find the investor that thinks that he needs to follow the crowd because they determine your profitability <b>is always the one wishing he had bought or sold at the lowest low or the highest high.</b> Simply put, I'm saying what I believe I have is a long term statistical edge in these newly created securities. The reason these ideas or results could not have been duplicated in the past is because QID, QLD, SDS, SSO and all of these other perfectly negatively correlated securities create the perfect environment to pairs trade on by the simple fact that their correlations are eternal. What correlations past Wall Street pairs trading models have found I would liken to spurious correlation that always dissipates. The reason these pairs models work is because you have a perfectly priceed, perfectly negatively correlated pair to arb every time on.
It will be impossible to arb daily closing prices out of the market. The only programs I know of capable of manipulating markets may only do so for very limited time periods, as much as 60 minutes, but no more, and they still cannot stop prices from moving where they need to go to find equilibrium.
So, I believe I'm doing cutting edge research here, and to doubt the significance of what I do you may ask only if you want to lose credibility.
Still working on the Cash Cow model. Light volume has prevented many trades from taking place, but that's part of the system.
Good Trading,
Good Luck,
Beau Wolinsky
To other matters.
I feel extremely confident in the results for Pairs Trading QID QLD 2.0.
I have been working towards applying the lessons learned from wash sales, autostops, and stop loss trades and have come up with a model for SDS and SSO. I also was working on FAZ and FAS, but it appears there is insufficient data to create a backtest with a statistically significant number of trades.
On the left is the SDS and SSO model, which I'm sure is obvious if you had followed my previous posts.
Long + Short Long + Short
Starting Capital $25,000.00 $25,000.00
Ending Capital $471,232.35 $1,081,660.08
Side by side, I can understand why someone would question the reason for the SDS and SSO lagging QID and QLD so much, but the annual returns show a different story, to me at least, having analyzed these reports for a long time.
Net Profit $446,232.35 $1,056,660.08
Net Profit % 1784.93% 4226.64%
Annualized Gain % 137.82% 203.89%
Exposure 74.79% 50.39%
Either one would still be a good system to use, till you see the DD.
Number of Trades 76 88
Avg Profit/Loss $5,871.48 $12,007.50
Avg Bars Held 4.91 2.8
Winning Trades 50 66
Winning % 65.79% 75.00%
Gross Profit $662,789.16 $1,727,862.95
Largest Winning Trades $74,588.57 $205,057.31
Avg Profit $13,255.78 $26,179.74
Avg Bars Held 5.08 2.86
Max Consecutive 7 10
A lot less trades with the SDS and SSO models. Hold times are also more than twice the QID and QLD version. Less of a win streak in the SDS and SSO models.
Losing Trades 26 22
Losing % 34.21% 25.00%
Gross Loss ($216,556.82) ($671,202.87)
Largest Losing Trade ($34,355.63) ($108,434.07)
Avg Loss ($8,329.11) ($30,509.22)
Avg Bars Held 4.58 2.59
Max Consecutive 4 2
As expected, a lot lower win percentage with SDS and SSO. Max consecutive is still inexcusable, but, by itself, not what I have identified as the cause of the Max DD.
Max Drawdown ($90,321.31) ($133,611.63)
Max Drawdown Date 8/11/2009 10/1/2009
Max Drawdown % -39.01% -20.34%
Max Drawdown % Date 10/8/2008 11/3/2008
So, here we are. Max DD for SDS and SSO is almost twice as much as the QID and QLD system. I attribute this to one trade, and one trade only. There is a trade in SDS and SSO that I can only call catastrophic. It is a losing trade that lost more than 17% in one day from 10/7/2008 to the Open on 10/8/2008.
Here is the tradeslist from that period. To show how volatility based overbought, oversold indicators adapt for changing market conditions, I have included the next two trades.
Position Symbol Shares Entry Date Entry Price Exit Date Exit Price % Change Net Profit Bars Held Profit/Bar Entry Signal Exit Signal Cum Profit MAE % MFE % ChartScript Delta Norm Series Actual Ratio Series
Long SSO 3,840 10/7/2008 41.53 10/8/2008 34.42 -17.13 -27,324.30 1 -27,324.30 Buy SSO Stop 29,766.54 -17.12 0.84
Long SSO 3,483 10/10/2008 27.09 10/13/2008 31.77 17.25 16,278.54 1 16,278.54 Buy SSO Profit Target 46,045.08 -11.04 19.23
Long SDS 1,538 10/21/2008 89.99 10/22/2008 100.47 11.63 16,102.24 1 16,102.24 Buy SDS Profit Target 62,147.32 -3.34 11.65
So, in defense of taking a loss like that, the next two trades were the number 1 and number 2 most profitable trades. I have attempted to find a way to "fit" this out, but mechanically, I have found no way and no other values to optimize the results any better, so I'm left to understand that to use SDS and SSO might warrant some possible attention to more than extreme volatility, and just plain irrational price action. I will need to ponder this for much longer, but the stats below give me pause as well.
APD 0.8301 0.7864
APAD 1.8742 1.7146
Oh the irony of having a system with 39% drawdown have a higher APD than a 20% drawdown. I can't make this stuff up. It's priceless, and a perfect example of what you'd go through to write 50 page long algorithms.
For those who don't know what APAD is, if APD is net profit over total drawdown, APAD is gross profit over total drawdown. The reason to look at APAD is to find whether your typical trade will overcome the total drawdown of the system, and I would say anything above 1 will.
Wealth-Lab Score 112.3872 322.2863
RAR 184.2791 404.595
MAR 3.5328 10.0224
Profit Factor 3.0606 2.5743
Recovery Factor 4.9405 7.9084
Before I discard the system, the returns below require some analysis. This WL score specifically might not accurately capture the underlying APR of the system as evidenced by the returns below.
Sharpe Ratio 1.8133 2.1658
My gold standard of high risk adjusted returns is always 2 on the Sharpe. Not that that will guarantee when you actually trade that you'll have a Sharpe of 2 the whole time, but that it has to be that high to <i>tolerate losses.</i>
Sortino Ratio 4.2993 5.4716
I have very rarely found any value to the Sortino ratio, because I truly believe backing out upside volatility masks upside drawdown. A true example of a system that has gamed the drawdown out of the equation through buy and hold is ETF timer. If you never sell a winner, you'll never have any drawdown, despite the position fluctuating 20-30k or 10-15% in a few days to weeks.
Ulcer Index 9.8026 6.1128
WL Error Term 29.287 8.755
WL Reward Ratio 4.7059 23.2885
Luck Coefficient 5.6269 7.8327
Pessimistic Rate of Return 2.1969 1.8607
Equity Drop Ratio 0.0508 0.0187
K-Ratio 0.3023 0.4032
Seykota Lake Ratio 0.0868 0.0449
Expectancy 0.6052 0.628
Expectancy Score 13.0956 15.9747
Max Losers Held 1 1
Max Winners Held 1 1
I don't find any value in the rest of these, but maybe someone out there will.
Now, to the part that I said I'd compare annual returns to. SDS and SSO is on the left, and QID and QLD is on the right
% Return % Max DD Period Starting % Return % Max DD
19.76 -6.22 7/13/2006 59.92 -10.19
-4.22 -28.55 1/3/2007 216.52 -12.33
370.79 -39.01 1/2/2008 237.8 -20.34
249.04 -24.5 1/2/2009 153.05 -16.23
So what does a professional see? I see a system that required a lot of data to be compiled before showing stellar returns. Every profitable system I've ever made has time periods of both stagnant and exponential growth. While I notice it took some time for the SDS and SSO version to come back, it has not benefited as much from the compounding effect the consistent returns generated by the QID and QLD system had, but that does not take away from the fact that the last two years the SDS and SSO version has significantly outperformed the QID and QLD version.
Now, I see confusion here. I know of quite a few funds that very long periods of time underperform, and in those few quarters of stellar growth essentially make the whole decade worth it for their investors.
I find trading to be a numbers game, with a lot more edge than card counting at Blackjack. Everyone will tell you to have a plan when you start, but they don't have any idea <b><i>how long it will take you to realize you need a better one.</i></b>
My approach to trading and investing is that I put myself from the perspective of passively buying an index, and benchmark myself to the S&P. By doing that, tolerating losses when the market goes down is much easier. In the periods where it goes back, I find myself always buying on the dip, which is another axiom prevalent in nearly every major trading site on the internet. It's certainly popular with professionals.
I liken the way I trade to a simple motto, and I know everyone has heard this before, but it's actually what I do with these pairs systems: <b>Buy when everyone's selling, and sell when everyone's buying.</b>
And that's all you have to do, no matter which timeframe your in. You'll find the investor that thinks that he needs to follow the crowd because they determine your profitability <b>is always the one wishing he had bought or sold at the lowest low or the highest high.</b> Simply put, I'm saying what I believe I have is a long term statistical edge in these newly created securities. The reason these ideas or results could not have been duplicated in the past is because QID, QLD, SDS, SSO and all of these other perfectly negatively correlated securities create the perfect environment to pairs trade on by the simple fact that their correlations are eternal. What correlations past Wall Street pairs trading models have found I would liken to spurious correlation that always dissipates. The reason these pairs models work is because you have a perfectly priceed, perfectly negatively correlated pair to arb every time on.
It will be impossible to arb daily closing prices out of the market. The only programs I know of capable of manipulating markets may only do so for very limited time periods, as much as 60 minutes, but no more, and they still cannot stop prices from moving where they need to go to find equilibrium.
So, I believe I'm doing cutting edge research here, and to doubt the significance of what I do you may ask only if you want to lose credibility.
Still working on the Cash Cow model. Light volume has prevented many trades from taking place, but that's part of the system.
Good Trading,
Good Luck,
Beau Wolinsky
It appears the ISM brought us down, and as I expected, chain store sales were up. The ISM happened to trigger a sell off that got me into SDS at 35.61 for Cash Cow. Unfortunately the market did not continue until later in the day and reversed about a half hour later. It sold on the reversal at 35.53.
What's significant to me about this trade is that it makes two consecutive trades in a row that have been completely automated. I think it is a matter of time before Cash Cow catches one of these reversals that lends to a trending day. In the backtest there were two day long trades that got in essentially on the first sign of either weakness or strength as the market continued in that direction.
I'm going to give Cash Cow a year to show what it can do. Most of the first six months were caught in technical problems out of my control, but such is the nature of the beast in implementation from theory.
QLD is on the oversold side, but not quite to an attractive enough level to take a position. The pairs model I have for SDS and SSO says SSO is a buy on the open tomorrow. I feel the system is a bit too risky, and I find QID and QLD is a sector I know much better than an amalgamation of various sectors.
What's significant to me about this trade is that it makes two consecutive trades in a row that have been completely automated. I think it is a matter of time before Cash Cow catches one of these reversals that lends to a trending day. In the backtest there were two day long trades that got in essentially on the first sign of either weakness or strength as the market continued in that direction.
I'm going to give Cash Cow a year to show what it can do. Most of the first six months were caught in technical problems out of my control, but such is the nature of the beast in implementation from theory.
QLD is on the oversold side, but not quite to an attractive enough level to take a position. The pairs model I have for SDS and SSO says SSO is a buy on the open tomorrow. I feel the system is a bit too risky, and I find QID and QLD is a sector I know much better than an amalgamation of various sectors.
Another trade in Cash Cow today, entered at noon exactly for SSO at 37.27 in C2, and at 37.29 for the model. Currently I show SSO at bid 37.75. I'm thinking this will be it's first big win.
I found the trade for the pairs model of SDS and SSO was correct in predicting the market's undervaluation, but it's entry was quite poor, and despite pegging the direction, since the entry was on the open, a very poor trade down 1.5% currently. I'm just watching that model currently and don't have any plans to trade or offer it.
Pairs Trading QID QLD 2.0 still shows us on the oversold side. Predicted values are as follows:
QLD Projection: 55.0869292709854 QLD Close: 55.1399993896484
QLD Projection: 54.4721817016602 QLD Close: 55.1399993896484
QLD Projection: 54.3628573826381 QLD Close: 55.1399993896484
Most likely we need to pull back to something less than 54.47 before considering a trade in QLD. Since we're not on the overbought side, I don't believe I need to include QID's predicted value.
I have done some research on the profit percent distribution of the trades and find the average profit accross all trades is 2.575%, with a standard deviation of 5.7665%.
This implies that 95% of the trades will be 2.575+/-5.7665*1.96=13.87734% or -5.53%. I believe it's good that 95% of the trades approximately already fall into this distribution, and especially the low value is a little beyond my stop of 5 and an 8th percent.
I also examined the kurtosis of the distribution and found that it is a less peaked distribution as evidenced by it's value being less than 1 at exactly .749182809. Many people would then conclude that that would imply there are fat tails possible in my system, but, you would have to look at the "skewness" to determine where those tails are, and, based on the skew of 0.927014752, I can conclude that negative values are quite limited compared to the huge profits on the right, positive side of the distribution. This is a good thing, and one day I hope somebody will realize just how good of a distribution that is.
I also, at some point, hope others would share their distributions with me, as I have, to compare. I believe there are tons of systems who may be exhibit lagging kurtosis with negatively skewed distributions that imply the system has "hidden risks" inherent in the system. A kurtosis below one, as I have said, means the distribution is "less peaked", and the step most forget then is to examine the skewness to determine "where the fat tails are at." In this case, if you had found a system with kurtosis below 1, and resoundingly positive skewness, you may conclude that the so-called "fat tails" are actually benefiting you in that they are "positively skewed, fat tails."
I encourage anyone to examine this distribution, and provide their current system for analysis. The point I'm making about "fat tails" is the proverbial "Black Swan" argument that really denies basic statistics. Certainly there are always outliers, but they don't happen very often. Once in a 1000 years even for some calculations of financial events, so the probability you hold it on that day is not even something to consider in your approach, and the "Black Swan" theory really has no basis in my opinion, because all it says is that <b>if there's always the possibility of a large move, you must not ever take risk</b>, which is a false assumption. Given that the probability of such events is so unlikely then if it does happen to you, you shouldn't change anything with what you were previously doing, as it can be considered economically and essentially a "sunk cost" so that it does not enter your strategic investment decisions.
% Change
16.09
15.88
15.88
15.87 avg 2.575227273
15.87 std dev 5.766540423
15.86 kurtosis 0.749182809
15.85 skew 0.927014752
15.83
15.81
11.31
7.88
7.73
6.51
6.49
6.22
6.19
6.15
5.69
5.32
5.16
5.11
4.84
4.5
4.42
3.78
3.57
3.35
3.01
2.96
2.93
2.84
2.57
2.5
2.41
2.35
2.24
2.23
2.16
2.11
2.08
2
1.88
1.77
1.72
1.7
1.7
1.66
1.64
1.61
1.61
1.41
1.4
1.37
1.35
1.34
1.32
1.13
1.13
1.02
0.94
0.88
0.72
0.67
0.63
0.52
0.24
-0.17
-0.48
-0.85
-1.21
-1.78
-1.88
-4.61
-4.62
-4.63
-4.63
-4.63
-4.64
-4.64
-4.67
-4.98
-5.14
-5.14
-5.14
-5.18
-5.25
-5.51
-6.51
I found the trade for the pairs model of SDS and SSO was correct in predicting the market's undervaluation, but it's entry was quite poor, and despite pegging the direction, since the entry was on the open, a very poor trade down 1.5% currently. I'm just watching that model currently and don't have any plans to trade or offer it.
Pairs Trading QID QLD 2.0 still shows us on the oversold side. Predicted values are as follows:
QLD Projection: 55.0869292709854 QLD Close: 55.1399993896484
QLD Projection: 54.4721817016602 QLD Close: 55.1399993896484
QLD Projection: 54.3628573826381 QLD Close: 55.1399993896484
Most likely we need to pull back to something less than 54.47 before considering a trade in QLD. Since we're not on the overbought side, I don't believe I need to include QID's predicted value.
I have done some research on the profit percent distribution of the trades and find the average profit accross all trades is 2.575%, with a standard deviation of 5.7665%.
This implies that 95% of the trades will be 2.575+/-5.7665*1.96=13.87734% or -5.53%. I believe it's good that 95% of the trades approximately already fall into this distribution, and especially the low value is a little beyond my stop of 5 and an 8th percent.
I also examined the kurtosis of the distribution and found that it is a less peaked distribution as evidenced by it's value being less than 1 at exactly .749182809. Many people would then conclude that that would imply there are fat tails possible in my system, but, you would have to look at the "skewness" to determine where those tails are, and, based on the skew of 0.927014752, I can conclude that negative values are quite limited compared to the huge profits on the right, positive side of the distribution. This is a good thing, and one day I hope somebody will realize just how good of a distribution that is.
I also, at some point, hope others would share their distributions with me, as I have, to compare. I believe there are tons of systems who may be exhibit lagging kurtosis with negatively skewed distributions that imply the system has "hidden risks" inherent in the system. A kurtosis below one, as I have said, means the distribution is "less peaked", and the step most forget then is to examine the skewness to determine "where the fat tails are at." In this case, if you had found a system with kurtosis below 1, and resoundingly positive skewness, you may conclude that the so-called "fat tails" are actually benefiting you in that they are "positively skewed, fat tails."
I encourage anyone to examine this distribution, and provide their current system for analysis. The point I'm making about "fat tails" is the proverbial "Black Swan" argument that really denies basic statistics. Certainly there are always outliers, but they don't happen very often. Once in a 1000 years even for some calculations of financial events, so the probability you hold it on that day is not even something to consider in your approach, and the "Black Swan" theory really has no basis in my opinion, because all it says is that <b>if there's always the possibility of a large move, you must not ever take risk</b>, which is a false assumption. Given that the probability of such events is so unlikely then if it does happen to you, you shouldn't change anything with what you were previously doing, as it can be considered economically and essentially a "sunk cost" so that it does not enter your strategic investment decisions.
% Change
16.09
15.88
15.88
15.87 avg 2.575227273
15.87 std dev 5.766540423
15.86 kurtosis 0.749182809
15.85 skew 0.927014752
15.83
15.81
11.31
7.88
7.73
6.51
6.49
6.22
6.19
6.15
5.69
5.32
5.16
5.11
4.84
4.5
4.42
3.78
3.57
3.35
3.01
2.96
2.93
2.84
2.57
2.5
2.41
2.35
2.24
2.23
2.16
2.11
2.08
2
1.88
1.77
1.72
1.7
1.7
1.66
1.64
1.61
1.61
1.41
1.4
1.37
1.35
1.34
1.32
1.13
1.13
1.02
0.94
0.88
0.72
0.67
0.63
0.52
0.24
-0.17
-0.48
-0.85
-1.21
-1.78
-1.88
-4.61
-4.62
-4.63
-4.63
-4.63
-4.64
-4.64
-4.67
-4.98
-5.14
-5.14
-5.14
-5.18
-5.25
-5.51
-6.51
- Status