Active Stock Trading

[target2019]

Rick Ferri put up a good blog post on the subject today:
Index Fund Portfolios Reign Superior

You can guess which side of the active-passive debate he's on...

I think the OP and subsequent posts are largely about active trading of individual stocks. The blog post is about the active/passive fund debate. He is very exuberant about his investing business and books. I prefer Bogle, he was "there" first!

 

[Nords]

I think the OP and subsequent posts are largely about active trading of individual stocks. The blog post is about the active/passive fund debate. He is very exuberant about his investing business and books. I prefer Bogle, he was "there" first!

If an active manager has a single-digit percentage hope of outperforming the market over 10-20 years, with all his professional training and advanced [-]insider trading[/-] analysis tools, then why should us retail investors be so naive as to think we can take advantage of the niches?

Yeah, sure, fund bloat. But individual stockpickers are reluctant to load up on individual shares when they should, and to hold on to them as long as they should.

 

[target2019]

The thread is about active trading. Wanted to hear more about that topic.

 

[brewer12345]

The thread is about active trading. Wanted to hear more about that topic.

Well, define active trading. Are we talking about systematized strategies, stuff driven by technical mumbo jumbo, or something closer to traditional active management? For example, I don't do rapid fire trading and technical analysis reminds me of the use of a dowsing rod, but I do trade in and out of equities, bonds and options based on valuation considerations and what I believe to be potential or actual price overshoots. I recently bought the longest tenor call options I could find on CORN because markets which exhibit very tight supplies, a supply shock, and relatively inelastic demand often wildly overshoot on the upside. No guarantees that this will happen in this case, but all the necessary preconditions exist and the options provide a way to get leveraged upside and limited downside.

 

[NW-Bound]

Compared to many people here who buy index funds and rebalance, or people who buy a balanced active fund like Wellesley or Wellington, I guess I can call myself an active investor.

I did not even know about the Bogleheads before I surfed this forum. I did not even know that my style was called slice-and-dice, so I could claim that I independently discovered this method.

As I started to experiment with buying individual stocks in the late 90s after rolling out my megacorp 401k after leaving it, and read different books on investment strategies, I observed that there were always hot and cold sectors. And I thought to myself that if one just rebalanced between these sectors, I could make a bit more money.

Of course, it would not help if one kept buying horse buggy sectors in the early 20th century, so it is not that simple. So, my performance has been a bit mixed. Some years I did well and beat the S&P500, some years not so. I take full responsibility for my action, but am reluctant to talk about it. It's simply an individual choice.

So, does the above make me an active investor? I do not do day trades, and can claim that my portfolio turnover is even lower than Wellesley, which is lower than most active funds. Of course, there have been times when I thought to myself that I would do a lot better if I did not hold on to my winners that long and if I traded more.

 

[Semiretired2008]

Myself, never really read any books. Read quite a few articles over the years (1.000's probably), but no books. Predominately over the years I have analyzed the markets, the methods that are used in determining key statistics and their uses, how individual stocks react to different circumstances and how this is viewed and used by market traders. With all of this I developed my own logical methods and developed my own thoughts on the subject. I continually adjust my methods as the markets change and as required for other reasons that arise.

If I would put a standard name to my style. I guess I would be a slice and dice, patterned volatility day - month trader so to speak for my active (frequent) trading style. For my dividend investments I am a value trader (value by my determinations - I use long term holds as long as they make sense). As I said in an earlier post - I am getting away from extremely high risk trading (long shots so to speak).

I use the standard key statistics that are available, but in a non-standard manner combined with the standard manner. I use what makes sense to me. Stock analyst use pre-built programs, spreadsheets, etc. with pre-built results - I do not believe in this - I do look at it, but do not let that guide my final decision. Over the years I have discovered weaknesses in just using straight forward stats.

I prefer to deduce my own decisions on a subject (not necessarily stocks - almost everything in life actually). Not everything fits into a standard mold. ER's are proof of that... I guess that is why I thought the subject would fit here.

When I started this thread I thought from what I had learned in this forum there would be a number of users interested in the topic - it made sense at the moment.

I truly believed that a bunch of INTJ's in retirement would be active stock traders (buying and selling at an above normal rate, but not necessarily high frequency traders). INTJ's are normally analytical, conceptual and have objective methods that they have developed in life that could be applied to the subject. I thought we could share the conceptual objectivity behind our methods. I can see that if I tried to post my methods the flak would be to high so have decided while I still like the subject I will keep the discussion of my methods at a minimum, but will join in on any discussion of the topic that others would like to post.

 

[NW-Bound]

Not all INTJs are alike when it comes to investing. Many look at statistics and see that the chance of beating index investing is not that great, and opt to go for the highest possible return for the least amount of work.

 

[rbmrtn]

I suppose I would be considered a "dirty market timer" by many as I use some technicals to exit/enter the market but I don't trade very often, not a day trader at all.

During the 90s I would have been a buy&hold, I didn't sell much because MrMarket didn't tell me to. I was still accumulating and 10+yrs away from retirement. Since 2000 I've made a couple of round trips. Last couple of years since leaving magacorp, I have focused on dividends for retirement income and capital preservation, and will continue this going forward. Current dividend payers cover expenses, I will add to the dividend payers as opportunities present themselves.

Nothing is 100%, B&H or timing.

There was a thread lasted year that got pretty hot, I think one poster got so frustrated he left, but I enjoyed the exchange of ideas.

http://www.early-retirement.org/forums/f28/market-timing-strategy-57014.html

 

[rescueme]

Not all INTJs are alike when it comes to investing. Many look at statistics and see that the chance of beating index investing is not that great, and opt to go for the highest possible return for the least amount of work.

Agreed.

As a (retired) INTJ, I would rather look at the norms and let somebody else to do the actual w*rk, while I/DW get on with the "business" of living.

As long as I (and my family) have more than we "need", we're happy.

Of course I/DW have already reached (and greatly exceeded) our "number", so we may have a different POV...

 

[NW-Bound]

I have always tried to curb my wants, as there are always things that other people have that I don't. As for needs, my needs are very elastic. I certainly can live on less than what I have now. So, why do I want to get more?

I look at investing and trying to get more returns as an intellectual exercise, and an exercise to overcome my greed and fear. It is also a hobby, not too different than what some people do as DIY projects.

When I no longer find it fun (hopefully that does not mean that I have lost too much ), I will put my money in a reliable MF like Wellesley or Wellington.

 

[EvrClrx311]

It is possible to beat the market consistently, but not by much and it does take a lot of work. I spent about 3 years reading books and researching forums on Mechanical Investing before ever putting my own money into the game. Currently I invest half of my taxable account in it... and leave retirement entirely to Index Funds.

My track record with active investing so far is good, but because I'm young and still don't have much skin in the game, the time involved hasn't been worth the added rate of return. I'd estimate I spend about 5-10 hours a week at it. I spent a good 500 hours a few years ago writing my own data miner and market analyzer to pull quotes and values from Yahoo Finance, it spits out 30 stocks once a month to invest in (I pay a broker $290 a year for unlimited trading). So essentially I'm taking emotion out of it entirely and trusting the numbers. History has shown that certain stocks that fit certain criteria (such as low P/E) tend to fall in the upper half of returns for a given time into the future. Pick enough of them together and you can edge the market average while staying diversified.

Over the last 7 years I've trailed the S&P500 once (by 4%) and beaten it once (by 14%)... the other 5 years I've edged it out by 1-6%. For me... it is more of a hobby... I've always loved programming, statistics and predictive modeling. I figure that if I can consistently beat the S&P by 3% a year on average in the long run it'll pay off... though I'm guessing I'll switch entirely to passive investing once I have enough involved to make me nervous.

Every active investor thinks they have a system that will work, even though 98% of others thought the same and failed. I'm trying my best not to be ignorant to that.

 

[photoguy]

History has shown that certain stocks that fit certain criteria (such as low P/E) tend to fall in the upper half of returns for a given time into the future. Pick enough of them together and you can edge the market average while staying diversified.

Over the last 7 years I've trailed the S&P500 once (by 4%) and beaten it once (by 14%)... the other 5 years I've edged it out by 1-6%. For me... it is more of a hobby... I've always loved programming, statistics and predictive modeling. I figure that if I can consistently beat the S&P by 3% a year on average in the long run it'll pay off... though I'm guessing I'll switch entirely to passive investing once I have enough involved to make me nervous.

EvrClrx -- Is your strategy different than simply buying into a value index? At least the way you describe it makes it sound very similar. Estimates of the value premium based on the Fama French models also seem to be around few percent (not sure exactly what the most recent research here shows) which would be consistent with your 3% goal.

 

[target2019]

When I started this thread I thought from what I had learned in this forum there would be a number of users interested in the topic - it made sense at the moment.

I truly believed that a bunch of INTJ's in retirement would be active stock traders (buying and selling at an above normal rate, but not necessarily high frequency traders). INTJ's are normally analytical, conceptual and have objective methods that they have developed in life that could be applied to the subject. I thought we could share the conceptual objectivity behind our methods. I can see that if I tried to post my methods the flak would be to high so have decided while I still like the subject I will keep the discussion of my methods at a minimum, but will join in on any discussion of the topic that others would like to post.

I see the wisdom of your initial post now, in testing the waters of discussion. I really do hope you post more on the topic. Perhaps the flak will be kept to a minimum.

As for concepts, I simply hold a diversified portfolio of mutual funds at this time. It's based on what I picked up from Bogle, Bernstein, etc. Only recently have I become interested in markets. On another forum I have become interested in tactical re-allocation and other methods to mitigate draw-down.

 

[EvrClrx311]

EvrClrx -- Is your strategy different than simply buying into a value index? At least the way you describe it makes it sound very similar. Estimates of the value premium based on the Fama French models also seem to be around few percent (not sure exactly what the most recent research here shows) which would be consistent with your 3% goal.

Its not quite as simple as Fama French, but uses the same principles. I subscribe to AAII and their SI Pro resources which release weekly updates on 9000 stocks covering 2,000 different data fields.

Way more data than a person can review on their own... but having that information is a fun challenge for computers. Track which stocks do best and then see which of those 2,000+ data fields that have in common... then try to understand why and how that group out performed the rest. Was it just a coincidence that Week? that Month? that Year?... and how likely are they to continue into the future.

Some patterns found are cyclical (I can tell you that the market over a 10 year period gains 50% of its total gains within a 4 day trading period each month... why? Maybe that's when people get their paychecks and invest most?)... others are temporary and as soon as you find them they go on to under perform (reversion back to their mean)

Humans are horrible at pattern recognition... we tend to find and make up correlations that don't exist. That is why I like to let the computer do most of the selecting itself... I just provide it the rules.

 

[Semiretired2008]

Its not quite as simple as Fama French, but uses the same principles. I subscribe to AAII and their SI Pro resources which release weekly updates on 9000 stocks covering 2,000 different data fields.

Way more data than a person can review on their own... but having that information is a fun challenge for computers. Track which stocks do best and then see which of those 2,000+ data fields that have in common... then try to understand why and how that group out performed the rest. Was it just a coincidence that Week? that Month? that Year?... and how likely are they to continue into the future.

Some patterns found are cyclical (I can tell you that the market over a 10 year period gains 50% of its total gains within a 4 day trading period each month... why? Maybe that's when people get their paychecks and invest most?)... others are temporary and as soon as you find them they go on to under perform (reversion back to their mean)

Humans are horrible at pattern recognition... we tend to find and make up correlations that don't exist. That is why I like to let the computer do most of the selecting itself... I just provide it the rules.

That AAII looks interesting. Currently I use a few sources for data to crunch in my tools.

How long have you been using AAII? Is the stream by minute or at least by 5 minute data downloadable for crunching locally?

I agree it is much easier to let the computer do it, but it still takes someone to tell the computer what to look for. I have been unable to find a reasonably priced tool that lets me alter and test rules at a whim and has download support. Fidelity's Active Trader Pro and Wealth Lab Pro offer some capabilities, but will not allow unique tweaks without major effort.

 

[Nords]

INTJ's are normally analytical, conceptual and have objective methods that they have developed in life that could be applied to the subject.

That's exactly why we aren't active stock traders. We've done all of that analysis, and our experiences have led us to conclude that the output is largely proportional to the amount of work input.

I have enough money to ER. I can't meet the criteria of the "Spend $10K!" thread. Why would I work harder to raise the size of the portfolio, especially when the downside risk is not insignificant?

As an INTJ ER, I'm happy to achieve 70% of the returns with about 10% of the labor. My "active" investing consists of (1) seeking startup companies for angel investments and (2) occasionally selling options on a stock or an ETF. Even then my experience indicates that I probably shouldn't be doing the former, and the latter is just part of forcing ourselves to rebalance.

 

[Semiretired2008]

That's exactly why we aren't active stock traders. We've done all of that analysis, and our experiences have led us to conclude that the output is largely proportional to the amount of work input.

I have enough money to ER. I can't meet the criteria of the "Spend $10K!" thread. Why would I work harder to raise the size of the portfolio, especially when the downside risk is not insignificant?

As an INTJ ER, I'm happy to achieve 70% of the returns with about 10% of the labor. My "active" investing consists of (1) seeking startup companies for angel investments and (2) occasionally selling options on a stock or an ETF. Even then my experience indicates that I probably shouldn't be doing the former, and the latter is just part of forcing ourselves to rebalance.

That is all fine and good - I do a lot of the same. I do not depend upon my success in it for my ER needs either - I would never count for its success to base my retirement on. For that I invest like everyone else - Mutual Funds, Stocks that pay dividends, etc.

For me it is something that I do that I also enjoy - I like the thought that it requires. I use the money I make with my active trading to pay for extra vacations and extra toys. If and when the time comes that I can no longer do it actively and or profitably then I will invest the money I use in this category into a Mutual Fund or something and go on with life, but for now it is something that I like to do and it keeps my mind active and for me that is important...

 

[photoguy]

Way more data than a person can review on their own... but having that information is a fun challenge for computers. Track which stocks do best and then see which of those 2,000+ data fields that have in common... then try to understand why and how that group out performed the rest. Was it just a coincidence that Week? that Month? that Year?... and how likely are they to continue into the future.

Ok. I have a good idea of what you are doing. In my other life, I basically spend my time building numerous statistical and other models out of noisy, crappy, & sparse data to make similar types of predictions (although not in stock market).

T
As an INTJ ER, I'm happy to achieve 70% of the returns with about 10% of the labor.

I haven't been following all of the most recent work in this area, but most of what I've seen suggests that there's no loss in expected returns for following a passive/indexing approach.

I agree it is much easier to let the computer do it, but it still takes someone to tell the computer what to look for. I have been unable to find a reasonably priced tool that lets me alter and test rules at a whim and has download support. Fidelity's Active Trader Pro and Wealth Lab Pro offer some capabilities, but will not allow unique tweaks without major effort.

If you have the inclination, try a statistical programming environment like R (it's free). Any sort of canned tool almost always has far too many limitations.

 

[Semiretired2008]

Ok. I have a good idea of what you are doing. In my other life, I basically spend my time building numerous statistical and other models out of noisy, crappy, & sparse data to make similar types of predictions (although not in stock market).

I haven't been following all of the most recent work in this area, but most of what I've seen suggests that there's no loss in expected returns for following a passive/indexing approach.

If you have the inclination, try a statistical programming environment like R (it's free). Any sort of canned tool almost always has far too many limitations.

That is kind of what I am doing. Getting live streaming data to my hard drive is the problem and or getting the logic I want in the canned tools is the other problem - they are all setup to use the standard type analysis methods - not home grown logic. I built predictive analytical type logic for a living for years. A lot of what I did was build logic that figured out other logic. Back in 2003 I started working to see what I could do with the stock market. I have done it on and off in my free time. On a small scale it works pretty good and since for me it is a hobby then small scale is fine. Fine tuning the logic keeps my mind active to boot...

 

[eatingmywords]

I actively trade, but consider my approach more cycle trading. I'm a product of the 90s and housing crash, so I take a very pessimistic view of the market. My view is the entire system is one big over-leveraged flame, so I try to ride bubble stocks up and get out before the crash.

Typically, I find high flying IBD stocks that have undergone significant corrections related more to the overall market behavior rather than the individual stock. I usually select stocks that had a prior rise of 30-50% in the last leg up and buy after a correction/dip (typically 2-3 months after peak). I get in early, let the market take the stock up, and generally get gains of at least 20%. Not all selections are winners, so i do use technical indicators for selling out early. This is very important in my system because the number of losers and the related losses would overshadow the winners.

There's obviously more than I'm willing to tell, but the amount of loss on any individual trade is generally less than 5%. After an initial rise, I will quickly move stop orders up to guarantee a profit (short of an algorithmic flash crash). I'm an INTJ / electrical engineer, and while I have no finance background, I do feel ashamed that with my signal processing experience my system isn't more sophisticated. I do recommend finding a system that can recognize cycle lows and highs and proving to yourself that these will work within reasonable probabilities (ie > 70%). I use relatively simple software, matlab for playing with ideas and tc2000 for implementing the system.

Company 401ks are a different matter, and I'm forced to hold for significantly longer periods of time. I use my wife's account as a control group for buy and hold in an age based retirement and compare to my account. I move money in my account at the bottom (or low) of a major cycle and out near the top. So far the cycle approach has twice the performance of the buy and hold approach.

 

[target2019]

Emw:
I am interested in how you define a cycle. Is this typically months, or perhaps shorter time frame? Or is it something driven by company specifics, so not easily applied to other companies?

 

[EvrClrx311]

That is kind of what I am doing. Getting live streaming data to my hard drive is the problem and or getting the logic I want in the canned tools is the other problem - they are all setup to use the standard type analysis methods - not home grown logic. I built predictive analytical type logic for a living for years. A lot of what I did was build logic that figured out other logic. Back in 2003 I started working to see what I could do with the stock market. I have done it on and off in my free time. On a small scale it works pretty good and since for me it is a hobby then small scale is fine. Fine tuning the logic keeps my mind active to boot...

My method of choice is transferring the data into SQL databases and crunching numbers and coding in PHP for a web based portal. C languages might be a little faster at the recursive and processor intense calculations... but I love how easy PHP was to pick up and how it interfaces with HTML...

I enjoy building websites in my spare time so its been a great way to keep up with the latest technologies. Also having the data accessible from the web means I can 'play' with my tools, even from my phone, if a new idea hits me while I'm well anywhere.

 

[photoguy]

My method of choice is transferring the data into SQL databases and crunching numbers and coding in PHP for a web based portal.

Very interesting.

Is there a stats package in PHP or other libraries to make analysis easy (besides a database connection)? How do you do basic model fitting or significance tests etc.? What do you do for graphics and visualization?

I've used PHP for my own websites but didn't know people used it for data analysis.

 

[pperl]

Regarding trading, while I don't usually get into returns since this apparently is an element of the discussion I will provide the following. As a long term holder my average holding period is 712 days or 1.95 years. The S&P Gain over this period was 25.81%, my average gain over this period, including dividends received, is 36.40% which is below my average gain over a longer period. As cited above, my focus is on commodity oriented equities-Agriculture, Energy, Precious Metals, Base Metals -- which comprise 79% of my investments, the balance being in banking, comp/tech, and miscellaneous. btw I am INFJ. All my best,

 

[EvrClrx311]

Very interesting.

Is there a stats package in PHP or other libraries to make analysis easy (besides a database connection)? How do you do basic model fitting or significance tests etc.? What do you do for graphics and visualization?

I've used PHP for my own websites but didn't know people used it for data analysis.

Yep, PHP has all of those things and I've used a lot of Google searching to find and learn how to use them

For Graphics I'm using the following free software (requires Flash 10 capable browser):
PHP/SWF Charts > Introduction

It's amazing the kinds of things you can do with it from pie charts to interactive graphs. Basically takes PHP code and data and turns it into Flash on the fly. A bit of a steep learning curve at first, but after 3 months playing around with it daily I became a semi-expert. Helps that I enjoy working with this stuff... otherwise it would not be fun.

For statistics and analysis there are a lot of great functions in PHP that take care of the busy stuff. I love all of the array functions that manipulate data quickly without having to write functions to do what you want to do (example: finding correlation coefficients, standard deviation, CAGR (Compound Annual Growth Rate), or Sharpe ratios is a single line of code most of the time.

I used to do a lot of the same modeling in C and although you have a little more control, it is A LOT more code and work to get to a result.