Corey’s Interview with Larry Connors Part 1

Oct 7, 2009: 10:30 AM CST

Recently, I had the opportunity to interview top trader Larry Connors, CEO and founder of TradingMarkets.com, and author of many popular trading books including Street Smarts (with Linda Raschke), High Probability ETF Trading, How Markets Really Work, and Short Term Strategies that Work. His full biography page (link) is listed at Trading Markets.com.

Mr. Connors is Managing Director of Connors Research, LLC, a private research firm that specializes in quantitative analysis for trading.  He has over 28 years of experience working in the financial markets industry.

His latest book, High Probability ETF Trading:  Seven Professional Strategies to Improve Your ETF Trading, was released in June, 2009 and is already in its second printing.

We discussed trading strategies, back-testing, quantitative trading strategies, Larry’s background and wisdom from experience, exchange traded funds, stop-loss logic, getting started in systematic trading, emotions and system trading, as well as what has changed from the past to the present, and what he sees for the future of trading.

I will be posting the full transcribed interview, divided into parts, over the next week.  This is part one of our discussion.  Thank you again to Larry for participating and for sharing his knowledge with us.

Corey Rosenbloom:  Thank you so much for taking the time to speak with me this afternoon.  Let’s talk a little about your background, your experience, and what got you interested in the markets in general.  What was the initial spark if interest and how did that develop over your life?

Larry Connors:  The passion was instilled early.  My grandfather had been in the stock market for many years as an investor.  As an 8 year old, he gave me a few stocks for my birthday – one of which was Exxon-Mobile – and from there, the interest continued to grow.  I traded in college and worked a couple of jobs in college, and over the summers, I worked in Fenway Park for the Red Socks as a vendor starting when I was 15-years old.  I used to use some of the money that was made from my jobs and found myself trading.

I went to work for Merrill Lynch shortly after finishing school at Syracuse University in early 1982 – at that time the Dow was about 800 or near 700 when I was hired.  A few months later, perhaps August, the “Great Bull Market” created by Reagan began.  Paul Volcker had cut interest rates and the Market was up and running.  Even though it’s had a big pullback over the last few years, it’s still much higher than it was back in 1982.

CR:  What really transitioned you from the Broker side of the market into the research and technical analysis into your passion for the research or quantitative side of the market?

Connors:  I was lucky enough to be working with some people at Merrill who understood technical analysis, and it was in the early stages of modern technical analysis and my understanding of it.  Beginning in 1987, I began taking it a lot more seriously, and I began with the idea that I was going to trade my account full-time.  At the time, there wasn’t a lot of information out there, which is hard to believe today, but back in the 1980s, the general consensus was that people who traded, did not share information.

So, the people who were writing books or putting any information out there were not doing it for a living, and you can see that from some of the things that were published back there.  Although it was not an easy path – it took me seven years to get there.  Ultimately, I got to that point in 1994 where I could go off on my own and trade for myself.  I had a private investment partnership with a few of my friends and we were profitable from every year forward that we ran it – it was good.

It is a very different psychology to go from making a good, nice living and knowing that money is always coming in then all the sudden, that source of income is cut-off and the sole source of income is just coming in directly from the markets.  It certainly focused my own trading, and if there was ever a time when I was most focused on my trading – it was when I first started out trading my own account full-time.

CR:  I read in a prior interview where you mentioned that you’re the best at your trading when you first start out on the floor (or as a new trader).  It’s definitely a struggle at first, but when you get initial success, complacency or overconfidence sets in and sometimes trading practices get sloppy.  Could you expand a little more on that quote and what it means to you?

Connors:  Sure – the story was told to me by Sheldon Natenberg who wrote the book Options Volatility and Pricing which has now of course become a classic.  The analogy that he gave was that “people trade at their best when they first come to the floor” because they were the most afraid of the markets.  Basically, they would just be trading small size and then they would have a little success and then they would slowly increase their size a little more and have more success but then they would get overconfident and ramp up their size and overleverage and then blow up and then he’d seen them the next week driving a taxi – they just blew up.

It’s probably truer now – a lot of people saw that happen in 2008, and ultimately as the stories come out that some of the over-leveraged bets that took some of these big firms down, like the type of leverage for example Lehman was reportedly taking on – it wasn’t bad trading that took them down, but it was the over-leveraged positions.

It’s probably true when it comes to FOREX trading – there’s a high turn-over rate amongst FOREX traders, and mainly it’s because it’s overleveraged.

The key is to understand how to use leverage – ultimately a great deal of wealth is created by using leverage, but leverage can also destroy wealth, so it has to be the fine balance of understanding how much risk one is willing to assume before taking on leveraged positions.

CR:  That’s one thing I’ve been picking up from comments on my blog and other sites as well.  A lot of people are drawn to leverage early in their career, for example, 3x Leveraged ETF Funds because they think they can make so much money in the shortest amount of time… but it just takes one mistake or bad trade on leverage to destroy wealth as quickly as it was made.  How can we foster new traders, or communicate to them the idea of “risk control” or “start small” to avoid the allure of quick, large profits with the risk of blowing out?

Connors:  Yeah, I think that has to start before – right from the beginning, where they have to make an agreement with themselves that they have to allow a certain amout of time to pass or reach a certain level of success before taking on any leverage.  People will come in to markets and just have a good run early on, and looking back to the market in 1999, it didn’t matter what you bought, you probably made a lot of money because the market was going higher.  Ultimately, it ended in the spring of 2000 of course.

Traders have to come in and say to themselves “Here, I’m going to start with $X amount of dollars” let’s say $100,000 “and this is the way I’m going to position size” and they may decide that they will not put more than 10% in any one position and not use any leverage and try to sustain the education and put in a process that puts a model into place that allows them to trade through multiple market cycles.  Unfortunately, it just takes time for multiple market cycles to unfold.

If we look at the run-up in the early to mid-1990s – it was a low volatility rising market, but as you get into the late 90’s, it was a high volatility rising market.  Different market that what was proceeded. In 2000 to 2002, we had a high volatility declining market.  2003-2007 – it was a rising market accompanied by declining or low volatility – in fact the VIX got under 10 for a few days there.  Finally, in 2008, we had a very high volatility, sharp declining market.

These are all different market cycles and one has to live through those. Whether one lives through those and does it bar by bar by hand – it gets them at least in the state of mind that markets will go through different stages both on a volatility and pricing basis.

This is exactly why we like to do our backtesting going back a couple of decades – or at least 15 years – we try to see all those different markets to get an understanding of where we are at any given time.

CR:  Let’s talk about backtesting – in your experience, what is the minimum size required in terms of years or trades for a statistically significant sample size for any kind of study an individual trader might look to do?

Connors:  Yes, it’s a great question – we used to test back to 1989 and cleaned the data.  The integrity of the data is always on the forefront of any backtesting method.  You want to make sure the data is clean and is accounting for the fact that some of the companies that are out of business now were in business back then, and taking account for splits and dividends.

What we started noticing is that the behavior of the market started changing in 1995, and part of that was the fact that the internet came into popularity of time, and CNBC and other financial shows started to be more widely watched, and information was being disseminated more equally in 1995.  As you moved forward into 1999 or 2000 – and of course now – if something happens that is newsworthy, the whole world – you and I – everyone is going to see it at the same time, but that wasn’t always the case.  That wasn’t true before 1995.

We like to go back and look at stocks at least back to 1995, but for ETFs, we do our testing since the first day of trading on each major ETF.  The SPY was the first one on the scene, which began in January 1993, so our ETF research begins there.  As each ETF came out, that’s the starting poitn of testing.

We now have a universe of over 700 ETFs, and we filtered it down to the top one-hundred or one-hundred and fifty ETFs based upon volume and we keep the leveraged and inverse ETFs out of the main universe.  We start with the first day of trading and compile it together.

I think anyone who wants to do any type of testing should do that.  If they’re trying to do testing on the leveraged ETFs – you mentioned the 3x ETFs earlier – and now those have a shelf-life of about one year, and that’s just not enough data to do any sort of backtesting.

CR:  Exactly.  The problem with those is that people treat them like options contracts it seems, and they deteriorate over time due to large percentage moves tied to the underlying index or fund.  People are emotionally drawn to those leveraged products for a variety of reasons, which leads us to a new topic.  In what sense can backtesting and strategy development eliminate or affect/minimize emotions and the effects of greed and fear?

Connors:  That’s a great question – the bottom line is that we’re all human.  Anyone who trades systematically or really in any type of trading wants to eliminate any type of emotion, and that’s normally not attained.  There’s always going to be emotion involved when something like money is at stake or at risk.

The nice thing about trading systematically or at least having models to guide you is that it does remove some of the emotion from the process if you allow the system to trade itself.  However, a lot of people tend to override models – the urge is always there to override models.  We have a position on right now, and the market is extremely overbought as we’re talking now, and our model has not triggered a sell signal yet and some of the people are wondering “why are we still hanging on to this?” but the model has not triggered a sell so we’re not going to override the model in something like that.

We’ve been doing this long enough to know that the ones (people) who are truly successful have to go through that process of saying “I’m going to be disciplined and let the model play itself out.”

People who have advantages at this are often people who have backgrounds in any type of rigorous science, mathematics, engineering – they tend to do better at model trading.  I see that in our customers too – when they come with that type of rigorous, non-emotional, scientific decision making background, it helps them and is certainly an advantage in trading.

CR:  What if the individual investor or new trader looking to apply systematic methods aren’t familiar with the scientific method or don’t have a research background?  How can they develop these skills to have a respect or knowledge of these quantitative methods?

Connors:  He could be gently scaled into the methods – let’s take an extreme example.  Say someone is a fully discretionary trader.  If he’s making money as a discretionary trader, there’s no reason to switch and become a systematic trader.  It’s only when he’s losing money consistently as a discretionary trader and he’s looking for a different method to adopt a different, more systematic strategy.

Ultimately, if the urge is there to trade on a discretionary basis, one can allocate account capital to both methods – say 50% to their discretionary methods and 50% to the new systematic basis.  That allows them to still trade on a discretionary basis and allows them to see which strategy works better as they let the system run its course over time.

Even to this day, we allocate 20% of our money (in our private investment partnership) to be able to trade things we call “Special Situations,” which are things we can’t quantify.  There’s behavior that we see but we can’t quantify but we want to take advantage of these opportunities that we haven’t rigorously tested.  We very rarely use that – it’s mostly in cash – but at least it’s allocated in the event an opportunity presents itself.  But if you look at the majority of trades we do in a year, it’s overwhelmingly based on the models.

Again, someone can gently bring themselves up to a higher percentage if they have success with the models if that’s what their goal is.

CR:  I like that approach better – it’s not just an ‘all in’ or ‘all out’ approach where they abandon what they’ve learned, which would be difficult to do.  Similar to what you mentioned in the High Probability ETF Trading book, that reminds me of the TPS (Time/Price Scaling Model) – that sounded like one of your favorite trade set-ups.  Here’s a two-part question – first, please describe that set-up as you defined it, and second… I’ve found that some of the trades that make us the most uncomfortable – like the TPS – tend to work the best, so if you could speak a little more about that notion in trading.

Connors: Absolutely – the TPS is hands down my favorite strategy – it allows us to scale into ETFs that are above the 200 day moving average which have pulled back, and allows us to get more aggressive in that scaling in process as it pulls back even further.  So, instead of going all in on a given pullback, we sort of dip our toes into the water, and if it pulls back further, we put on a little more position and get more aggressive.  Ultimately, what we’re doing is getting to a full position later.  We’re not doubling up or putting on leverage, we’re just averaging into a full position a little bit at a time.

The downside to that is that you have to have high cash levels to use that approach – basically cash levels are usually high to do so.  But historically, it’s high probability trading as you can see from the book – and we’ve expanded and detailed the TPS Strategy.  We show the backtested models and take a look at the ETFs going back to the inception of trading – you’ll see results anywhere from the mid-80s to the low 90s in terms of percent of correct or winning trades using TPS and the type of scaling model.

We really like that, and we’ve moved into using TPS in stocks, and we now have about a dozen stock trading systems or strategies and we’re having internal discussions on how to incorporate TPS into those, or how the TPS method encompasses those dozen stock systems and we’re continually running tests on that.

So, it’s a nice way for people to be able to trade.  We’ve seen good success from people who have been using it.  I have a number of people who subscribe to my daily subscription service called “The Daily Battle Plan” and the model portfolio focuses on the TPS Strategy on ETFs, and we scale into ETFs and people have had great success over the past year since we introduced the model.

CR:  A lot of the research on higher probability trading results -even in the research I’ve done – the trades themselves sometimes look very scary to enter.  But it almost feels that the trades that work the best are the ones that are the hardest to enter – like perhaps being in a deeply overbought or oversold territory.  Talk a little bit about the battle or balance – really across all trading strategies – about the need to feel good/safe psychologically  (which often doesn’t lead to profitable trades) and the need to make money (which often leads to trades that make us feel terrified at the time).

Connors:  Yes, it’s a good point – if you take a look at the times we’re buying… there’s sometimes no other way to describe the charts other than “gross.”  The worse the charts look, usually the bigger the edge is.

When we’re putting on positions – for us, it’s just symbols – I don’t even look at the charts anymore – I know what the charts look like, and in the few times I look at the charts, I’m just like everyone else and wonder “why are we buying these stocks?  What the heck are we doing?”

Especially when volatility increases,when you get into these high volatility times.  August 1987 was one of our better months, and I have to tell you – it was scary!  It was scary doing some of the positions we were putting on.  But in hindsight, it turned out to be one of the better months because those charts were pretty ugly, but that’s where the edges are. It’s where most people won’t tread.

If everybody is playing in the same area and looking to pick up money – it’s not there.

CR:  It’s a zero-sum game, right?

Connors:  It’s pretty close, and in fact, in some cases in a number of the methodologies that are common and public out there in regards to some of our testing, we show that people are going out there with negative edge, meaning that they’re going in and buying positions that have negative edge, and that’s of course not where you want to start.  You want to start off of course with a positive edge.

If too many people are playing the same way, then obviously the edges are gone.

If you take a look at a number of books out there, they all say the same thing.  When you start putting statistics to it to quantify it, you see that there’s no edges, and in many cases, there’s actually negative edges.

CR:  Do you factor in the volatility measures, perhaps by using the ADX or Bollinger Bands or the VIX or some type of indicator to assess the volatility environment of these ’scary’ positions?

Connors:  Yeah, we actually want volatility for trading.  The inverse is true for investing, but we want volatility for trading.  We usually put volatility filters to say that we want historical volatility to be at least 30% – in fact the higher the better.  We need things to move.  We don’t want to be in things that are moving sideways – we need our stocks to have range.

A friend of mine runs a pretty large prop fund, and he says, “You’re a volatility trader,” and I never viewed it that way, but that’s basically what we are.  The higher the volatility, the better for us.

________________

Stay tuned for part two of our interview.

  • Digg
  • del.icio.us
  • Facebook
  • E-mail this story to a friend!
  • StumbleUpon
  • Technorati
  • TwitThis
Comments
  • alcorwin

    We are great fans of Larry Connors and own several of his books, but we are puzzled by TPS. We have tried over a thousand different tests, and TPS seems to lose to the all-in strategy ALWAYS. At first we thought we must be doing something wrong because so many of Larry's assertions hold up well in the face of the most intense scrutiny, but our current conclusion is that TPS keeps too much of your money on the sideline too long.

    We have searched for statistics that Connors might have based his assertion on, but so have not found those statistics. There may be right there somewhere in plain sight. They usually present solid statistical evidence, and it's just a little hard to find this evidence.

    We did find two statistics that could make someone think that this was working. The second and third levels bets do have higher win percentages and greater return on investment than the original bets, but that never overcomes the costs of keeping money in reserve to average in.

    All-in sounds like a dangerous strategy, and certainly you need to exercise external money management so it is only your stake in this tournament that is at risk. Once you have bankrolled one of Connors strategies, you will maximize profitability by following an all-in strategy.

    Making this easier to swallow is the knowledge that with almost every set of parameter settings makes a gross profit in the last twenty years, and the parameters that Connors recommends have shown a worst case max draw down of four percent in the SPY in that time. All-in doesn't put anywhere near 100% of your money at risk. It does tie up all of your money, however, and that can be just as bad in some cases.

  • Gary

    Awesome interview so far. Can't wait for the next installment. Thanks!

  • Rob L

    This is my problem with a lot of their (trading markets) strategies. The charts are friggin SCARY. It works terrific if the market pulls back and then has a big up day, their picks tend to out perform the market 3x... but you're trying to pick the bottom, which is never a great idea (or as Larry called them 'gross' charts). I signed up for the service for a month and during that time I made a TS scan that almost mirrored their picks (trading well above the 200 SMA, 2 period RSI under 10)... not exactly magic.

    The strategies that Corey teaches work much better and I MUCH perfer the feeling of understanding a chart compared to poking at GROSS charts of oversold stocks that should out-perform the market (that doesn't mean it won't continue to spiral down, just maybe less than the S&P).

blog comments powered by Disqus