A Nasdaq 100 Component ATR Trailing Stop System

Forecast page: https://www.tradingtimemachine.com/trading_forecast.html

The forecast gauge remain much like the market…chop. Wake me up whe. it hits red or green.

Thanks for reading Trading Time Machine! Subscribe for free to receive new posts and support my work.

---

Tonight, I’d like to share another backtest with you using all the components of the Nasdaq 100.

This is a system that beginning backtesters often ask me about. They have a TradingView backtest with questionable historical data, a very limited data set, and they fail to account for survivorship bias by removing symbols that have been delisted from the index and adding new ones on their appropriate addition dates.

High quality, clean datasets account for these changes. For example, in the backtest I’m going to show you, the current symbol list contains 101 symbols representing 100 companies, Alphabet has two classes of shares: GOOG and GOOGL. Since 1985 at the start of the index, more than 300 symbols have been part of the index. That is a lot of companies not being accounted for. Many of the current members of that Naz 100 were not around at the start of your testing period.

They often point to simple moving average crossover systems, looking at some of the impressive winners in recent years, and begin to believe that the decent results they’ve seen are indicative of real-world performance.

They are not.

I’ve often seen this method shared on YouTube and other various social outlets. It is always very selective in the symbols they showcase.

The System:

The Entry

We will use the survivorship bias removed, clean Nasdaq 100 dataset.

We will enter when the 50 period moving average crosses the 200 day moving average.

The portfolio will allow for 10 holdings (10% of the index).

The Exit

We will apply an ATR based trailing stop to catch the big winners and get out of losers.

The value I see thrown around frequently is a 4 ATR Trailing Stop (based on the recent 14 bars)

Pretty straightforward.

You’ll see some nice winners like:

ROST recently

Or INTC recently

Nice trades. The trailing stop did its job and caught nice runners.

But what if we look at all of the trades?

What are the statistics of that trading portfolio?

APR _ 6.75%

Maximum Drawdown _ 52.79%

Win % _ 42.34

Average Winner 14.19% - Average Bars Held _ 34.24

Average Loser -7.13% - Average Bars Held _ 14.32

Profit Factor _ 1.41

Sharpe Ratio _ 0.52 (this tells us things are going to be pretty wiggly relative to our APR.

Here are your monthly and annual returns

Here is the plot of portfolio value versus Nasdaq 100 in white

Kind of bumpy like the index right?

Maybe we just didn’t have the right settings. Maybe a 4 ATR is too tight. Maybe the 14 day lookback for the ATR is not long enough. This system would probably benefit from a longer holding period by widening that ATR trailing stop. Our winners only held for 34 bars.

We should run an optimization testing all variables.

We will look at every ATR trail level from 1.0 to 8.0 - stepping in increments of 0.2. So we’ll look at 1.0, 1.2, 1.4, 1.6 and so on up to 8.0.

We will also test every ATR period from 4 days to 80 days - stepping up by 4 each time. 4, 8, 12, 16 and so on up to 80.

When optimizing this that creates 720 possible combinations - for example a 6.2 ATR with a 24 day lookback.

So what are the best combinations?

Generally when analyzing with things like Sharpe ratio or APR ,using an ATR about 6 has worked best. The lookback periods didn’t actually make a big difference.

If we select the best scenario - 6.0 ATR with the 36 Period the results are as follows.

APR _ 13.43%

Maximum Drawdown _ 48.99%

Win % _ 46.57

Average Winner 26.00% - Average Bars Held - 126.84

Average Loser _ -10.25% - Average Bars Held - 48.74

Profit Factor _ 2.22

Sharpe Ratio _ 0.77

Those are all huge improvements over the initial settings.

Having that wider stop was much more beneficial in letting the trend develop.

The monthly and annual returns.

The Equity Curve

Not horrible. It has had a good run from the 08’ lows. It is a good trend following method with the typical stats and results you’d expect from a system like this.

Look how it stayed in this AVGO trade recently. Very nice.

But….Think about the people that are going to attempt something like this. Do they know the size of the drawdowns historically? Do they know the expected returns?

Here we can see the drawdown levels of the system in red compared to the Naz 100 index in white. Those are some pretty big drops relative to the annual return.

Without a complete picture, they may think something is wrong when their portfolio hasn’t reached a new high in two years, when that would actually be normal. Psychologically, by not knowing the expected path, they will be quick to abandon the system and try something new, typically after a bad run. Once they switch to something new, it will take time to get that portfolio back to its previous highs or potentially to new lows. Not unheard of.

I have seen this over and over with people who have come to me through the years. They want me to either test a set of rules they have or test what they have been doing for recent months or years.

Then I test. Reality hits them hard.

Be very careful of systems or methods that have not been thoroughly tested.

This next statement is going to be a big one - but it is true.

I have tested every defined system on every timeframe from every book, post, or video.

All of them.

It is what I have been doing for the last 30 years. It is what I do every day and many nights (my wife will attest).

Very very few public systems have an actual portfolio path that was even close to what was represented. Most are complete BS. It is something the reader or viewer tries, and then they move on after the first challenge to their belief system.

Again, I’ve seen it MANY times.

Good systems are rare. Even those good systems have holes that could benefit from another system that does well when that system does poorly.

Think about the base system of buy and hold. It is a very good method to drive increasing wealth over time. A mix of Index funds and a core Bond portfolio goes a long way.

Those bonds can temper those really bad periods for equities.

The hard part is sticking with it. Sustainable wealth creation comes from leaving it alone and not letting the occasional rough patch in the markets sway you. That is not easy. Those 30% portfolio drawdowns can be gut wrenching for investors.

I love to build systems that help those people. I try to build simple to execute strategies that can make the path easier to stomach, which allows the investor to stay engaged and stick with that long term wealth building.

My Elite SPY Trading System is the flagship of the systems I offer to the public. It is off the charts good.

I have a new system coming in the next few days. This one is going to do specifically one thing. Trade when the VIX is high. It generates positive returns when most portfolios or strategies are taking hits.

This will be a single lifetime price thing. You would get all the rules, the backtest data, every historical trade, and spreadsheet that calculates the signals for you. It’s pretty cool and would provide serious utility to most people.

I’m considering offering an early access version where Substack readers could get it at a heavily discounted price, and in return, provide feedback to help me work out any issues or suggest ideas to be integrated.

Anyway, I’m the worst salesman ever. I hate doing this kind of thing, but at times I feel I have to push it a bit because I genuinely believe it could be a great benefit to most people. Online finance is full of seriously sketchy content. What I build is not that. I’m always transparent and honest with my readers and clients.

In my opinion that goes a long way .

Have a Great Night!

Dave Johnson - Quantitative System Designer of the Elite SPY Trading System

TradingTimeMachine.com

Thanks for reading Trading Time Machine! Subscribe for free to receive new posts and support my work.

via Trading Time Machine https://ift.tt/hP5VEsl