Charles Kirk Q&A with Larry Connors: Quantitative Trading and Holy Grails, Part 3

Today, we have part 3 of our Charles Kirk’s extensive interview with Larry Connors (Click here to read Part 1 and Part 2).

Charles Kirk is the founder of The Kirk Report, one of the more widely-read websites on trading and the financial markets available to the average retail trader. As Larry has noted, this interview is one of the most thorough interviews Larry has ever granted and we think Battle Plan subscribers will enjoy and benefit from the ideas and insights from this conversation.

If you have any questions or thoughts about the interview, feel free to e-mail Larry with them.

Charles Kirk Q&A with Larry Connors, Part 3

Courtesy of The Kirk Report.

Charles Kirk: Are there any well-known strategies that worked really well in the past that you believe do not work in today’s market?

Larry Connors: I’ve always been fascinated by the Turtle methodology of trend trading and how it’s evolved. I’m not a breakout trader. But I certainly have a great deal of respect for the Turtles and many others who do trade that type of trend following strategy in the futures markets. A lot of wealth has been created from trend following strategies. John Henry was able to accumulate enough wealth using a trend following methodology to purchase a Major League baseball team (the Boston Red Sox).

Kirk: Some believe that even an excellent strategy that has been backtested can be ineffective in the hands of a trader who doesn’t follow its signals or can’t control their emotion, has this been experienced as well?

Connors: Yes. I mentioned that before and it’s absolutely true.

Kirk: Now that we’ve talked about the concerned quantitative strategies, let’s now talk about the benefits. You’ve written a couple really good books lately.

You’ve also written about what you call the “Holy Grail” of indicators. What is that indicator and how do you use it?

Connors: Thank you for the kind words, Charles. I’m glad you enjoyed the books. I consider the 2-period RSI (Relative Strength Index) to be the single best indicator that’s available for traders out there. This is for equity traders, especially equity and ETF traders. We started publishing research on the 2-period RSI as far back as 2003, but the RSI was created by Wells Wilder. He deserves all the credit. Back in the 1970s, Welles Wilder created what was originally a 14-period RSI and if you take a look at most charting packages they default to that 14-period RSI.

To get a 2-period RSI, just change the period from 14 to 2.

We combine that with the 200-day moving average. In fact, I can give you three simple rules that can be used on 150 actively traded, non-leveraged ETFs since the inception of each.

Through the end of May 2009, with an ETF that’s trading above its 200-day moving average, buy when its 2-period RSI closes below 5. And exit when its 2-period RSI closes above 70. On the short side, if the ETF is below the 200-day moving average, short when its 2-period RSI is above 95 and exit when its 2-period RSI closes under 30.

Over a couple thousand trades, involving 150 ETFs, the per trade win rate comes out to a little bit over 75 percent, with an average edge of 1.3%. You’ll see it’s a strategy that has been consistently profitable for many years.

What is the 2-period RSI doing? With the 200-day moving average as a filter, the 2-day RSI lets traders buy into weakness when it goes below 5 and it’s selling into strength when it goes above 70. Again, it starts with the same philosophy we trade over and over again; buy into weakness and sell into strength. The 2-period RSI probably takes advantage of that philosophy better than any other indicator that is out there.

Kirk: Buying pullbacks is something you find helpful. Can you give me some examples of recent trades using a 2-period RSI that worked more recently?

Connors: If you want to see many examples of this philosophy in action, you can take a free trial to the Daily Battle Plan, my daily trading service that you can access at the TradingMarkets website. There’s a model perfotlio that has been running now for a year. You’ll be able to see the setups and be able to punch up the trades and you’ll see how they unfolded. It’s good and so far, in 2009, the majority of the trades in the model portfolio have been successful. So those are good examples for someone to be able to follow and be able to watch.

Kirk: I know you discuss other great methods in the book and I recommend those interested to check them out. In your view, what has been the most powerful trading strategy from the point of view of the feedback you have received?

Connors: On the ETFs side, the strategy was called TPS, which stands for time, price, and scale-in. The initial research was published in High Probability ETF Trading.

It’s a very simple strategy that uses the 2-period RSI and the 200-day moving average and scales-in to positions. TPS is the strategy that I use, for example, to guide me in taking trades in my daily trading service, The Daily Battle Plan. It is the primary strategy that we use in our own trading for ETFs. I feel it’s the single best strategy that we’ve created and published to date.

Kirk: Are there any situations where you go against a trade that starts moving below its 200-day moving average on the long side?

Connors: Not often, and the few times that it would be done would be examples of what I called before “special situations.” We’ve allocated a certain amount of capital for that kind of trading, for trades that we haven’t quantified.

We also have a TPS strategy that scales-in to extremely oversold ETFs that have historically been high probability trades.

Kirk: Many professional traders will tell you that knowing when to exit is as important as knowing when to take a position. You outline using the RSI as an effective exit strategy. But you don’t talk about other exits like trailing stops. What is your thinking on exits?

Connors: We use dynamic exits. A dynamic exit is an exit that is adjusting to price. By comparison, a static exit would be buying a stock at $50 and deciding to exit five days later, or exiting at $52. Those are static exits.

We prefer dynamic exits, exits that move with price, a 2-period RSI or a cross beyond a 5-day moving average. In our testing, dynamic exits tend to improve results compared to static exits.

As far as trailing stops and fixed time exits, no matter what we’ve done, it’s hurt the performance of many, many models we’ve created. That’s been the case with any type of stops that have been added to the models.

We think that stops are a form of insurance. But one thing traders tend to forget is that stops do absolutely nothing to protect from overnight risk. So if you buy a stock at 50 and it opens the next day at 30 – and that certainly does happen – stops will do nothing to protect you.

We feel that position sizing is a better way to be able to limit risk. This way, no individual position can cause too much harm to a portfolio. We also feel that using certain options strategies are better and more efficient in limiting risk than stops.

Kirk: As someone who sees the value of utilizing several different types of strategies, based upon the different indicators, how do you manage many different strategies, providing conflicting signals?

Connors: We don’t often get conflicting signals. That doesn’t happen, where one strategy says a market is overbought and another says that it is oversold.

But to help your readers, one of the largest quant fund managers in the world told me that he has many, many hundreds of models. He has models that do give conflicting signals. What he does is to take all signals, so basically he can be long the market and short the market. The reason he does that is that he wants to see how the model performs in the real world. So even though they’ll be net flat a position, they are still isolating the model by itself and able to see how the model performs at the end of the year.

For us, though, we don’t often see conflicting signals. We’re buying into the selling and selling into the buying.

Kirk: Do you utilize different time frames in your analysis?

Connors: The majority of our research and the majority of our trading are tied into short-term. So that means from 3 to 7 days on average on a hold. We do have some capital allocated to a day trading strategy as I mentioned before. Also, if you go to the Connors Research web site you’ll see that we do have research for longer term investing at www.connorsresearch.com. For example, we have quantified that the volatility of a stock is basically a good indication of how that stock will perform over the next 252 days. It’s a good research study. You can find it at connorsresearch.com.

Kirk: Most of your strategies you’ve shared show how to manage the trade entry and exit, but now how do you find attractive trades in the first place? Can you talk briefly about how you scan for new opportunities?

Connors: We have that all built in. We use platforms, multiple platforms, in order to be able to trigger the signals, depending upon which of the strategies and the type of time frame that we use.

Kirk: In your most recent book, you stated that ETFs are safer than common stocks because they don’t have the same corporate risk. Are you now suggesting an individual focus strictly in ETFs?

Connors: I’m not suggesting that they strictly focus on ETFs. ETFs are wonderful vehicles and we’ve allocated more of our capital to ETFs because of the opportunities that we think are there. But that takes nothing away from stocks, it takes nothing away from options. We still do trade those.

ETFs, though, really provide traders with so many opportunities and so many different ways to grow your money. They are a wonderful innovation. In my opinion, they are the best innovation since options were created. And I think there are many professionals that would probably agree with that.

You can see from the volume that has gone into ETFs that they’ve really capture the trader’s imagination. The volume is not coming from investors, it is coming from traders who are using them and really taking advantage.

Kirk: Let’s talk about your latest book, High Probability ETF Trading. First why so much focus on the 200-day moving average?

Connors: We’ve tested all sorts of moving averages and taking nothing away from the other moving averages like the 50-day, which is very popular. What we have found is the 200-day moving average, by itself, tends to be the single best moving average to help guide whether or not to be invested long or short side.

All one has to do is look at 2008. Had all these mutual fund money managers and hedge fund managers simply avoided stocks after they broke under the 200-day moving average, you can see how much wealthier people would be today. That’s simply by avoiding those stocks below the 200-day moving average.

The 200-day moving average tends to be where we draw the line and the test results continue to hold up years after we originally published the research.

Kirk: Talk about the performance results in the ETF strategies in the book. Were slippage and commission costs factored in? How would that have changed the results?

Connors: We focused the book on the 20 most liquid ETFs at the time we started writing the book. So, they were the 1x ETFs. We also avoided having one sector factored in more than once.

Including slippage and commission would have minimal impact on the test results.

Kirk: In the book, you talk about having a preference to trade country fund ETFs over sector and commodity bond funds. What’s the reason behind this?

Connors: The backtesting.

There are different types of baskets. A basket of all technology stocks will reduce single stock risk. But there will still be sector risk. If Intel misses their earnings numbers, for example, the entire Semiconductor basket will end up ultimately moving with Intel. So there’s less single stock or corporate risk, but still sector risk.

When you get into country ETFs (baskets of stocks from a single country), there tends to be even less risk Outside of Iceland, countries just don’t go out of business. They can be more volatile as a whole – take a look at FXI which is China or EWZ for Brazil. But as a whole, when you take a look at the test results, what you’ll see over and over again is that the country funds tend to produce overall better reversion to the mean test results.

Kirk: You said that traders must be aware of so-called correlation risk and in fact you said that was the reason so many money merchants failed in 2008.

Simply put what do some traders need to know about managing correlation risk and portfolios?

Connors: If you took a look at February as the U.S. went and there was no such thing as diversification.

I know people will go out there and look at correlations over three years, four years and five years. I don’t believe that’s correct. Correlations are constantly changing. It’s a dynamic process. So, we’ll look at a 21-trading day correlation. If you take a look at many of the ETFs out there, you’ll see very high correlations, especially in February or March 2009. So it didn’t make a difference which position you are in; they were all basically performing the same.

With that being said, now that it’s come into June of 2009, things have gotten a little bit more back to normal. We’re seeing correlations now going back to normal.

So, one needs to be aware of when things are correlating, and that’s why a dynamic correlation is built in to help. Again, with a 21-day look back, a one-month look back we’ll tend to give you a better idea of what the correlation currently is.

Courtesy of The Kirk Report.

Click here to read Part 4 of the interview with Larry Connors.

Larry Connors is CEO and Founder of TradingMarkets.com and Connors Research. His two most recent books are Short-Term Strategies That Work and High Probability ETF Trading. Larry also has a daily ETF subscription service. For a free one-week trial, click here.