High Probability ETF Trading by Larry Connors and Cesar Alvarez

1) Introduction

ETFs have become popular with traders for several reasons. They have lower risk compared to individual stocks, they cover all the major and many minor sectors, and some offer leverage or inverse trading. There are many ETFs with enough liquidity to be a good source for short or long-term trades.

• 20 of the highest volume ETFs were selected for testing the strategies. Their ticker symbols are: DIA, EEM, EFA, EWH, EWJ, EWT, EWZ, FXJ, GLD, ILF, IWM, IYR, QQQQ, SPY, XHB, XLB, XLE, XLF, XLI, and XLV.
• The ETFs were tested from January 1993, or when the ETF started trading, through December 31, 2008. These were simulated trades without allowing for commissions or slippage. Neither leveraged nor inverse ETFs were tested.
• The strategies have the same basic concept. Buy the ETFs only on pullbacks and short them only on short-term strength.
• The ETFs were bought only when above their 200-day simple moving average and shorted when below their 200-day moving average.
• Exits were made dynamic by adjusting to the current price.
• The testing showed country ETFs moved between overbought and oversold better than sector ETFs. And sector ETFs tended to trade between overbought and oversold better than individual stocks.

2) The 3-Day High/Low Method

Identifying when a security is overbought or oversold is a common objective for traders. The closing price is usually the reference to indicate if an extreme price move has occurred. The 3-Day High/Low Method relies on the daily highs and lows, instead of the closing price.

1. The ETF today closes above its 200-day simple moving average and below its 5-day simple moving average.
2. For today and the two previous days, each day's high and low is less than the prior day's high and low.
3. Buy on today's close.
4. If prices close below the initial entry price during the trade, buying a second unit is an aggressive option.
5. Sell when the ETF closes above its 5-day simple moving average.

The 20 ETFs were tested from their inception to 12/31/08

3-Day High/Low Results:
Long Basic Version

	Trades	Ave P/L %	Win %
Lowest	4	-0.29	64.7
Highest	64	1.79	88.9
Median	36	0.64	74.7
SPY	64	0.91	87.5

If a second unit was added below the initial entry price, the winners and average profit to loss improved.

3-Day High/Low Results:
Long Aggressive Version

	Trades	Ave P/L %	Win %
Lowest	4	0.10	70.6
Highest	64	2.14	91.7
Median	36	0.95	82.2
SPY	64	1.12	90.6

1. The ETF today closes below its 200-day simple moving average and above its 5-day simple moving average.
2. For today and the two previous days, each day's high and low is above the prior day's high and low.
3. Sell on today's close.
4. If prices close above the initial entry price during the trade, selling a second unit is an aggressive option.
5. Cover when the ETF closes below its 5-day simple moving average.

3-Day High/Low Results:
Short Basic Version

	Trades	Ave P/L %	Win %
Lowest	2	-0.23	50.0
Highest	38	3.63	84.4
Median	23	0.84	70.7
SPY	32	0.67	84.4

When a second unit was added, most win rates improved, as did the median, and profit to loss. The regular short tests had three ETFs with losses, but with the aggressive option version, they were all profitable.

3-Day High/Low Results:
Short Aggressive Version

	Trades	Ave P/L %	Win %
Lowest	2	0.09	50.0
Highest	38	3.43	91.7
Median	23	1.30	78.6
SPY	32	0.89	84.4

Like the other strategies in the book, the 3-Day High/Low Method is a simple way to find overbought or oversold ETFs. Using the 200 day moving average as the long-term trend indicator, this type of short-term counter-trend method can be very effective.

3) RSI 25 & RSI 75

The buy side of this strategy was created in 2003. Since its publication in 2004, it has performed well. It was designed for trading SPY and QQQQ, since those were the most popular ETFs at that time. The number of ETFs has greatly increased since then, and this method continues to be profitable for most of them. The short side was added to the method shortly before the book's publication.

The method uses a modified version of Welles Wilder's Relative Strength Index (RSI). The indicator is designed to measure the price trend strength of a security. It gives a single number from 0 to 100 with the strongest trends indicated at the high end of the scale.

RSI was created with a default period of 14 days. This method replaces the 14-day period with a 4-day period, RSI(4). The change was made to give different indicator readings compared to the regular RSI. This modification provides an edge over the majority of traders who use the standard 14-day period.

1. The ETF is above its 200-day simple moving average.
2. Buy on the close if the RSI(4) closes under 25.
3. If RSI(4) closes under 20 during the trade, it's an aggressive option to add a second unit.
4. Sell on a RSI(4) close over 55.

The 20 ETFs were tested from their inception to 12/31/08

RSI 25 Results:
Long Basic Version

	Trades	Ave P/L %	Win %
Lowest	4	-0.03	64.4
Highest	84	2.11	86.1
Median	39	1.00	75.8
SPY	84	1.23	82.1

Adding a second unit with the aggressive option improved the median, profit to loss, and the win percentage.

RSI 25 Results:
Long Aggressive Version

	Trades	Ave P/L %	Win %
Lowest	4	0.45	73.9
Highest	84	2.62	89.3
Median	39	1.48	81.3
SPY	84	1.48	89.3

1. The ETF is below its 200-day average.
2. Short when the RSI(4) closes above 75.
3. A second unit can be shorted if the RSI(4) closes over 80. This is optional.
4. Cover the short on a RSI(4) close under 45.

There were relatively few short trades for this method. The result was a wide range of win rates. The rates varied from 50.0 percent to 100 percent (on 3 trades for one ETF and 5 trades for another). The profit to loss was equally inconsistent from -0.92 (on 5 trades) to 8.93 (on 3 trades).

RSI 75 Results:
Short Basic Version

	Trades	Ave P/L %	Win %
Lowest	4	-0.92	50.0
Highest	42	8.93	100.0
Median	21	1.39	69.8
SPY	30	0.68	63.3

Adding the optional second unit improved the lowest win percentages and profit to loss averages, which showed up in higher medians.

RSI 75 Results:
Short Aggressive Version

	Trades	Ave P/L %	Win %
Lowest	4	0.03	54.5
Highest	42	9.09	100.0
Median	21	1.84	77.3
SPY	30	0.94	70.0

The RSI 25 and RSI 75 Strategy was originally created for use with only two ETFs. Yet despite that, it has proven to work well with most other ETFs since it was first published.

4) R3 Strategy

Every trader needs an edge to be successful. One way to get a small edge, is to modify a technical indicator from the default settings. The usual setting for the RSI indicator is 14 periods. As with the previous method, a shorter period was used for the R3 Strategy. This strategy uses only a 2-day period, RSI(2). This trade setup came out in 2003 and as the tests show, continued to work well years after it was first published.

1. The ETF is above its 200-day simple moving average.
2. The RSI(2) goes lower for the previous 2 consecutive days with the first of these down days going below 60.
3. Buy on today's close if the RSI(2) is below 10 and below the previous day's RSI(2).
4. While in the trade, if prices close below the initial entry price, the aggressive option is to add another position.
5. On a RSI(2) close over 70, exit the trade.

The 20 ETFs were tested from their inception to 12/31/08

R3 Results:
Long Basic Version

	Trades	Ave P/L %	Win %
Lowest	5	0.17	60.0
Highest	73	1.94	84.9
Median	36	0.90	76.4
SPY	73	1.16	84.9

Adding another unit made good improvements to the average profit to loss ratings and a moderate increase in the win rates.

R3 Results:
Long Aggressive Version

	Trades	Ave P/L %	Win %
Lowest	5	0.36	69.8
Highest	73	2.48	89.0
Median	36	1.20	80.0
SPY	73	1.38	89.0

1. The ETF is below its 200-day simple moving average.
2. The RSI(2) goes higher for the previous 2 consecutive days with the first of these up days going above 40.
3. Sell on today's close if the RSI(2) is above 90 and above the previous day's RSI(2).
4. While in the trade, if prices close above the initial entry price, the aggressive option is to add another position.
5. On a RSI(2) close under 30, exit the trade.

R3 Results:
Short Basic Version

	Trades	Ave P/L %	Win %
Lowest	3	-0.62	0.0
Highest	40	4.46	100.0
Median	20	1.37	71.5
SPY	34	1.13	70.6

The aggressive version reduced the losers from 4 to 1 (on 3 trades) and the median value improved.

R3 Results:
Short Aggressive Version

	Trades	Ave P/L %	Win %
Lowest	3	-0.33	33.3
Highest	40	5.37	100.0
Median	20	1.64	76.5
SPY	34	1.31	76.5

Despite using only 2 days to signal a trade, the 2-period RSI is very effective in spotting short counter-trend moves.

5) The %b Strategy

Bollinger Bands, a popular technical indicator, was created by John Bollinger. A feature of this tool, is a measure called %b. It produces a single number to describe the location of the current price in relation to the bands. %b indicates how low or high current prices are compared to recent prices. Typically, long trades are made when %b is low. When %b is high, it's a possible selling or shorting opportunity.

1. The ETF is over its 200-day simple moving average.
2. If %b closes below 0.20 for 3 consecutive days, buy it on the third day's close.
3. If during the trade, %b closes below 0.20, the aggressive option is to add a second unit.
4. When %b closes above 0.8, exit the trade.

The 20 ETFs were tested from their inception to 12/31/08

%b Results:
Long Basic Version

	Trades	Ave P/L %	Win %
Lowest	7	0.18	68.2
Highest	111	1.92	87.9
Median	51	0.67	74.5
SPY	111	0.79	82.0

All the ETFs were profitable in the basic and aggressive versions. Adding the second unit with the optional aggressive version increased all metrics.

%b Results:
Long Aggressive Version

	Trades	Ave P/L %	Win %
Lowest	7	0.23	69.7
Highest	111	2.17	93.9
Median	51	0.88	80.5
SPY	111	0.95	85.6

1. The ETF is under its 200-day simple moving average.
2. If %b closes above 0.80 for 3 consecutive days, sell it on the third day's close.
3. If during the trade, %b closes over 0.80, the aggressive option is to short a second unit.
4. When %b closes below 0.20, exit the trade.

%b Results:
Short Basic Version

	Trades	Ave P/L %	Win %
Lowest	6	-0.18	56.7
Highest	67	4.63	100.0
Median	33	0.90	69.5
SPY	48	0.74	77.1

There was one losing ETF with the basic version. It remained the worst performer with the aggressive version, but did show a profit.

%b Results:
Short Aggressive Version

	Trades	Ave P/L %	Win %
Lowest	6	0.14	57.1
Highest	67	4.75	100.0
Median	33	1.24	75.0
SPY	48	0.98	79.2

This strategy was created in 2004 for use with SPY and QQQQ. As with the other methods, this strategy has continued to show profits with most major ETFs. With only three simple rules, there were 20 out of 20 winners on the long side. Reversing the rules produced 19 winners out of the 20 ETFs, for the basic short sale version. The aggressive version had all 20 as winners.

6) Multiple Days Up (MDU) and Multiple Days Down Strategy (MDD)

How can you tell when a security goes too far against its major trend? That's what this and the other strategies try to answer. In this case, the Multiple Days Up and Multiple Days Down Strategy, looks for ETFs which have had mostly lower or higher closes for several days. Multiple lower days when an ETF is above its 200-day moving average, offers a possible long trade. Likewise, multiple higher closes when an ETF is below its 200-day average, is a potential short.

1. The ETF is higher than its 200-day simple moving average.
2. The ETF has closed lower in 4 of the last 5 days.
3. If the ETF closes below its 5-day simple moving average, buy on the close.
4. While in the trade, the aggressive option is to add another unit if prices close lower than the initial entry price.
5. When the ETF closes higher than its 5-day simple moving average, exit the trade.

The 20 ETFs were tested from their inception to 12/31/08

MDD Results:
Long Basic Version

	Trades	Ave P/L %	Win %
Lowest	7	-0.17	56.5
Highest	110	1.45	88.0
Median	53	0.49	73.2
SPY	110	0.50	74.5

The basic version had rather ordinary results and one losing ETF. But the aggressive version showed strong improvement and no losers.

MDD Results:
Long Aggressive Version

	Trades	Ave P/L %	Win %
Lowest	7	0.10	67.4
Highest	110	2.03	92.0
Median	53	0.84	78.6
SPY	110	0.78	85.5

1. The ETF is below its 200-day simple moving average.
2. The ETF has closed higher in 4 of the last 5 days.
3. If the ETF closes above its 5-day simple moving average, sell short on the close.
4. While in the trade, the aggressive option is to short another unit if prices close above the initial entry price.
5. When the ETF closes below its 5-day simple moving average, exit the trade.

MDU Results:
Short Basic Version

	Trades	Ave P/L %	Win %
Lowest	4	-0.69	25.0
Highest	56	1.99	80.0
Median	37	0.79	70.7
SPY	50	0.88	80.0

There was one ETF which lost with both the basic and aggressive versions. But it had only 4 trades, which may not reflect its long-term results.

MDU Results:
Short Aggressive Version

	Trades	Ave P/L %	Win %
Lowest	4	-0.15	25.0
Highest	56	2.47	85.4
Median	37	1.3	77.8
SPY	50	1.13	80.0

This method shows you don't need complex rules or computer-generated indicators to come up with a profitable strategy. Two moving averages and counting a few closes can be all you need. This strategy, like any of the others, can be used as a starting point for a similar method. MDU and MDD can be applied to other strings of up or down days, such as 4 out of 6 or 5 out of 6 days.

7) RSI 10/6 & RSI 90/94 Strategy

This strategy shows another way to profit using a modified version of the Relative Strength Index (RSI). Like the R3 Strategy, this method uses a 2-day period RSI(2) for an edge over the default 14 days. Since this strategy requires extreme RSI(2) values to signal a trade, it produces fewer trades than some of the other methods. But the overall results can make the wait worth it.

1. The ETF is over its 200-day simple moving average.
2. Buy on the close if the RSI(2) drops below 10.
3. During the trade, add another unit if the RSI(2) closes below 6.
4. When the ETF closes over its 5-day simple moving average, exit the trade.

The 20 ETFs were tested from their inception to 12/31/08

RSI 10/6 Results:
Long

	Trades	Ave P/L %	Win %
Lowest	6	-0.61	66.7
Highest	117	1.93	88.7
Median	57	0.88	80.5
SPY	117	1.04	86.3

There was one ETF showing a loss over 6 trades. All the others were profitable with 20 trades or more.

1. The ETF is below its 200-day simple moving average.
2. Sell short on the close if the RSI(2) is above 90.
3. During the trade, short another unit if the RSI(2) closes over 94.
4. When the ETF closes below its 5-day simple moving average, exit the trade.

This short strategy was very successful with no ETFs showing a loss and good overall results.

RSI 90/94 Results:
Short

	Trades	Ave P/L %	Win %
Lowest	5	0.38	40.0
Highest	66	5.55	86.8
Median	34	1.47	78.3
SPY	47	1.30	76.6

Individual ETFs have this strategy setup about 6 to 8 times a year. Trading a variety of ETFs will result in several times those numbers. Like the other strategies, the values used are not the only ones capable of showing a profit. More extreme values such as 5 and 2 instead of 10 and 6, will reduce trade numbers, but may increase overall profits. Simply test and use what works best for you.

8) An Introduction to TPS^℠

TPS^℠ is the newest strategy in the book and the most powerful. TPS^℠ stands for Time—Price—Scale-in. Using these three elements produces a strategy with the highest winning percentage of any strategy the authors have created. TPS^℠ takes a position in an overbought or oversold ETF. Then additional positions are added as the ETF becomes more overbought or oversold.

The original version of this strategy was taught to a five-figure a year membership group. Since its first publication in 2008, there have been tens of thousands of variations taught to the membership group. Naturally, only the basic version is shown here, without all the various refinements possible.

1. The ETF is over its 200-day simple moving average.
2. For two consecutive days the 2-day RSI is under 25. Buy 10 percent of your full position on the second day's close.
3. Make the following additions only as long as the ETF remains above its 200-day simple moving average.
4. During the trade, if prices close below your initial entry price (step 2), buy 20 percent more of your full position.
5. During the trade, if prices close below your last entry price (step 4), buy 30 percent more of your full position.
6. During the trade, if prices close below your last entry price (step 5), buy 40 percent more of your full position. If you reached this point, you have 100 percent of your full position
7. When the 2-day RSI closes above 70, exit the trade.

The 20 ETFs were tested from their inception to 12/31/08

TPS^℠ 1-2-3-4 Results:
Long

	Trades	Ave P/L %	Win %
Lowest	5	0.82	80.0
Highest	133	2.97	95.7
Median	59	1.34	88.6
SPY	133	1.27	93.9

Even the ETF with only 5 trades won on 4 of the 5 and had an average profit to loss of 1.13. The others were equally or more successful.

1. The ETF is under its 200-day simple moving average.
2. For two consecutive days the 2-day RSI is over 75. Short 10 percent of your full position on the second day's close.
3. Make the following additions only as long as the ETF remains below its 200-day simple moving average.
4. During the trade, if prices close above your initial entry price (step 2), short 20 percent more of your full position.
5. During the trade, if prices close above your last entry price (step 4), short 30 percent more of your full position.
6. During the trade, if prices close above your last entry price (step 5), short 40 percent more of your full position. If you reached this point, you have 100 percent of your full position
7. When the 2-day RSI closes under 30, exit the trade.

As with the long trades, all the ETFs made profits going short. They had excellent win percentages and average profit to loss ratios.

TPS^℠ 1-2-3-4 Results:
Short

	Trades	Ave P/L %	Win %
Lowest	6	0.82	37.5
Highest	69	6.63	88.0
Median	34	2.04	81.9
SPY	55	1.49	85.4

These rules for TPS^℠ are only the basic ones and can act as a foundation for thousands of profitable variations. You can use different RSI levels, change the number of days for a signal, and adjust the way you scale-in a position. Since there are thousands of profitable alternatives to the values presented here, you can test and find out what works best for you.

9) Think Deeper

The book, The Talent Code, describes a process which can be applied when using the strategies in this book. The author, Daniel Coyle, says talent is developed through Deep Practice, not by your genes.

Malcolm Gladwell's book Outliers, popularized the idea expertise is developed after 10,000 hours of deliberate practice. Coyle says there is a level beyond deliberate practice, which he calls Deep Practice.

Deep Practice requires three steps. The first step is Chunking It Up. Break up all the parts of a talent which need to be mastered, into smaller pieces. Then practice each of these individual talent chunks.

To apply chunking to the book's trades, requires dividing the strategies into two basic parts. The long-term trend, up or down, as determined by the 200-day moving average is the first part. The second part is unique to each strategy. The strategies use different ways to show an extreme reaction to the major trend. With one way to measure the major trend, and multiple ways to measure the counter-trend, you can move on to the second step of Deep Practice.

The Deep Practice second step requires in the zone type of concentration. Mindless repetition won't do, only fully engaged practice will make the trading talent a part of you.

The final step of the process is, Learn to Feel It. You not only think about what you practice, but you feel it too. Consider what your thoughts and feelings are when you practice a trade.

Paper trade a strategy and concentrate on your reactions from trade entry to exit. Your practice needs to reflect real trading situations with good and bad outcomes. Notice the difference between trades which end quickly and those which take more time. Trades entered or exited on the close lose potential intra-day profits. Be aware if the thought of lost additional profits is an issue for you. Do this process correctly enough times and you're on your way to trading mastery.

10) Putting the High Probability ETF Trading Pieces Together

The book's seven strategies have shown the benefit of buying and shorting minor trend reactions. When an ETF is above its 200-day moving average, it's a time to look for buying opportunities. And look for possible shorts when it's under the average.

ETFs were exclusively chosen for these strategies because they are more predictable than stocks. Unlike stocks, they don't have the volatility caused by earnings reports and other company news. ETFs won't have the same rapid advances as a hot tech stock, but also won't drop like a rock on a corporate scandal.

Strategy Questions

Do these strategies work with leveraged and inverse ETFs?
The strategies were only tested on 20 liquid regular ETFs. Leveraged and inverse ETFs perform differently than regular ETFs. If you want to use leveraged or inverse ETFs, you should test them first to see how their performance compares.

Why were stops not included with the strategies?
If you can place stops and get good results, then use them. But stops were not included for the simple reason they are an expensive form of insurance. They don't improve profitability. With this type of strategy, stops tend to exit a trade just before prices turn around for a profit. And stops don't eliminate the risk of carrying a trade overnight. Position sizing and hedging with options offers a better way to manage risk.

How should I choose an ETF to trade when many are overbought or oversold on the same day?
There are a number of ways to make a choice, but there are three simple options. The first, is to divide the ETFs by category and rank them in the following order: country, sector, commodity, currency, and bond fund ETFs. Country ETFs have the best test results and tend to be safer. Another way to choose is to go with the ETF with the best RSI rating. When buying, you want the lowest reading and when shorting, the highest. The third way is to look at the historical record. Since some ETFs have performed better than others, you can choose the ETF with the most consistent test results.

Can I use ETF options?
Although not specifically tested, the strategies should work well with options. ETFs with high win rates can be a good fit for credit and ratio spreads. As always, test before you trade.

What should I know about the risk of correlation?
Correlation is an important concern when trading ETFs. Correlation describes how related securities tend to move in the same direction. ETFs in the same category usually will be overbought or oversold at the same time. This happens on most days with index ETFs which follow the Dow, S&P 500, and the Nasdaq. By choosing more than one ETF in the same or related category, you may think you're diversifying, but actually, you're only increasing your risk.

What more do I need to know about trading ETFs?
You need the discipline to follow a plan you created. When the news is telling you one thing and a strategy is opposite to the news, it's hard to take the trade. But if you want to be successful, you must block out the noise and follow your plan. You need to decide what strategies you'll use and how you'll apply them. What ETFs will you trade? How will you decide position size and other risk control methods? Will options or leveraged ETFs be a part of your plan? Preparation and trade execution will determine your success.

Will these strategies work for stocks?
Although many of the ideas behind the strategies came from stock trading, stocks present risks which ETFs don't have. The strategies have shown they work well with ETFs and that remains their intended use.

What if I want more information about ETFs?
There are many sources, but when this was published, there were three sites you might try. IndexUniverse.com gives a good overview of the ETF world. SeekingAlpha.com aggregates blogs about ETFs. ETFTrends.com can help keep you up on industry information.

Glossary

All the basic terms used with the strategies from %b to Relative Strength Index are described.