There are basically two approaches to directional trading of E-minis: i.) hit and run based on technical patterns, and ii.) trend trading with a mechanical system.
The Afternoon Strategy
During the afternoon, we use a mechanical strategy to catch trends. This 89 MB is best viewed over Wi-Fi and may take several minutes to buffer over a slow connection.
The Theory
I conducted two webinars at eSignal to explain why we do what we do.
The description for the hour-long presentation was, “Trading E-mini stock index futures is more fun and profitable when traders break the day into two parts. Hit and run technical trading tactics tend to work well during price discovery in the first hour while a mechanical, trend-trading strategy captures the big up or down days that often emerge in the afternoon. In this webinar, Teresa reveals the rationale behind her personal trading strategy for intraday E-mini S&P trading, complete with data analysis, backtests and conclusions.”
Q&A
A new member had some questions about SmarterSuite for TradeStation after reading about SmarterStops in The Very Latest E-Mini Trading, 2nd Edition: Using Market Anticipation to Trade Electronic Futures by Michael Gutmann.
JF
Thank you. I have watched the webinar from 18 August, and found it very helpful.
What I am looking for on the smarter stops is an explanation of the thinking behind it and how the user-defined preferences are designed to influence the output. The quickstart guide does not appear to provide this, but I may be just looking in all the wrong places. In reality, what I am trying to identify is whether the stops are going to be a good addition to my trading style.
Also just as feedback, I am finding the website a little difficult to navigate. I feel as though I have not read some sort of crucial introduction that helps me navigate it sensibly and find what I am looking for. If you have any advice on how a newbie to your site should structure their reading, I would be grateful.
Teresa
The site has accumulated a lot of information over the past few years, so yes, there is way too much information. One suggestion is to scan the headlines in the Knowledge Base to see if any of them seem relevant to what you’re looking for.
As for SmarterStops, I originally made them so that people could *test for themselves* what happens if they trade when it’s volatile vs. not volatile, what happens if they employ high vs. low leverage, what happens if they use tight vs. loose stops.
Ultimately, it turned out that nobody cared, and everyone just trades tiny time frames with too much leverage with super tight stops when it’s really volatile. That’s the recipe for disaster if you ask me, but I cannot convince them otherwise. So now, as you learned from the eSignal webinar, I apply it the way I think it should be and broadcast the signals on the ES so they can see what I personally do with these stops.
As for how they are constructed, I wrote a bit here but as they are proprietary, I can’t say much other than the calculation does not involve averages and is based on proper application of statistical methods.
The common ATR stops require people to guess just how many ATRs they should use, but I didn’t like that sort of discretion, hence “universal”. I basically engineered a proper volatility stop by getting rid of the use of averages and other incorrect statistical “methods” employed by others.
Since you have TradeStation, you can also incorporate the functions into your own strategy.
JF
Thank you, yes I saw that piece, and largely agree with your comments. Volatility requires wider stops and smaller positions.
I understand what you are saying, but does that suggest you should not fool with the factory settings on the indicator – i.e. just run it and let it be? It is important to be clear because there appears to be the ability to manipulate user defined variables in the TS version.
Teresa
The point was to allow the users to verify my suggestions for themselves rather than take my word for it, but as I said, it turned out that nobody was willing to do the work and want me to just tell them what to do.
My suggestion is you should do the same and test the various parameters so that you can feel confident when you use the system.
JF
Yes I know, and I have a real intellectual issue with those sorts of inputs into my trading.
The other thing I wanted to mention was (clearly because of lookback period periods) the stops change if you use the 24 hr globex chart or the day session. I have always traded only the day session, but would you have any comment on this? It levels out later in the day, obviously when the lookback periods are dealing with the same data.
Teresa
I think if you are using a 24 hour chart, you should use tick charts because time charts would reflect low-volume trading overnight, and therefore, the price is like an auction with a handful of people in the room. Or at least start at the same time as the bond market. As you saw in the webinar, I do not use the stops in the first hour of trading.
JF
The other thing is, among your indicators, which gives the best measure of volatility? What I would like to do is create a measure of volatility for the first hour of the trading day and compare it against a “norm”.
Teresa
Volatility can only be estimated, and so you can use any number of measures. The indicator that is included in SmarterSuite is called Smarter.Range and it’s in percentage so you can do comparisons across instruments.
JF
I have been going through the knowledge base as suggested, and from what I gather the setting that you discuss moving is the one that varies between between 1.0 and 1.5. What is the impact of this, ie what does moving this variable do?
Teresa
Smaller = tighter stops.
JF
There appear to be four user inputs in the TS file, and I guess what I am trying to get to is an understanding of what the inputs control. If what you are saying is it’s trial and error and up to me, that’s cool, but I don’t want to unnecessarily re-invent the wheel.
Teresa
Again, those were added so people would not have to take my word for anything.
They can test leverage by adjusting OwnCapital/PercentOwnCapital. I advocate low leverage but they use high leverage and lose money.
They can test tight/loose stops with Universal Stops or set it themselves using the StopFactor. I advocate stops not too tight but they use tight stops and lose money.
They can test for themselves if they should use the same position size when shorting. I advocate 1/2 position size when short because “down” = higher volatility, but they short the same size and lose money.
They can test for themselves if they should use a moving average with the TrendFilter. I say it doesn’t help, but they use them and get whipsawed.
They can test for themselves if they should trade all day with TradeStart and TradeEnd. I say don’t do it, but then do and wonder why they lose money.
They can test for themselves if they should trade when volatility is high with RangeFilter. I say stay away from high volatility but they go for it and lose money.
So in fact, this is a way for those who cannot write code to test the most commonly held ideas.
JF
I also note that intra-bar breaching of the stop does not seem to reverse the signal (like Welles Wilder’s Parabolic stop). Is your system designed to exit only if the stop is breached on close, any breach, or not defined?
Teresa
The quick start guide says:
Apply the indicator (Insert > Indicator > _Smarter.Stops) to a chart with a couple of clicks. Use the dots plotted by _Smarter.Stops as trailing stops according to the following two rules:
- BUY RULE: A *close* above a blue dot is buy signal.
- SELL RULE: A *close* below a pink dot is a sell signal.
JF
I take your point on 24 hour charts, and don’t actually use them for trading (only reviewing overnight action before the day starts) – I trade off the 5 min day session chart, but there is no reason why running the stops off a tick chart is not possible.
This discussion highlights to me how uncomfortable I have become with black boxes.
Teresa
Again, SmarterStrategy was NOT intended as black box. It is a way for people who cannot write code to test some common approaches. Only the calculation of the stop is proprietary and even then, it was constructed using classic univariate time series techniques appropriate for the stylized facts of asset price returns.
Users can learn why they have been losing money without risking their money. If you wish to write you own system and incorporate only the stops like Mike Gutmann, you can do so with the functions that same with SmarterSuite. There is no need to worry about all the other parameters if you’ve done your homework and they don’t apply to you.
JF
This is an entertaining insight, as I spent years fiddling with ADX and Adapative Moving Averages in all sorts of weird and wonderful ways. I love how trading teaches you about yourself.
Teresa
I guess the problem is that ADX and averages might not be that useful in trading in general, and you learn that so much of trading is misapplied univariate time series analysis.
Could you repeat that please
In English? ;-)
I thought that your English was perfect…(laugh)
I found your comments on ADX and other indicators interesting because it seems like you used to use the ADX and DI+ and DI- quite a bit in your trading. Before 2005?
I read your Guide to Discretionary trade set-ups and most of the charts have either ADX or DI+ and DI- on them.
Do you think that the indicator “Stopped working” or do you think that the indicator was fundamentally flawed from the beginning?
Do you see any benefit to using these tools now in your guide to disgressionary trade set-ups?
Based on my own testing: DI+ and DI- still seems useful.
I used those indicators because they were there, before I learned the time series analysis techniques. These tools still have value; they will get the job done, but the ones I built since are an improvement in terms of calculation.
Hi Teresa,
I just wanted to comment on your afternoon trading system. I have found it helpful, for me, to see the afternoon trading, as if, I was considering sailing a boat & that the volatility levels as the winds available to me for sailing. Low volatility, low or little wind, not good for sailing. The higher the winds the better the sailing. The thought can to me, is their a volatility threshold, so to speak, that is equivalent to a storm warming— sailing not advised. Am I just barking at the wind or is their something to my feeble attempt at analogy?
Yes, you are right to say there is a low level where it’s just “dead”, but too high is also bad, except in intraday trading, we’ve done away with “too windy” with the use of tick charts.