Market Timing based on Portfolio returns

Is there any way in the “Hedge Market Timing” entry/exit rules to set up your portfolio to go into cash when your port folio has dropped 10% and then get back in the market when it has gained back the 10%?

The general idea is to base the market timing on portfolio performance instead of individual stock performance.

Regards,

John.

John
Can’t be done I’m afraid.
This is how I do my ‘market timing’ but I have to use other software to do it.
Even just adding a few technical indicators to the interactive graph of the equity curve would be better than nothing.
Steve

If your portfolio goes to cash, there’s no way for it to gain 10%. You can time it based on an etf like SPY.

You can export a simulation equity curve to excel and then apply equity curve timing in excel to determine how it performs in a backtest. For a portfolio you could just let the portfolio run and adjust your actual trading accordingly.

J:

 If you know what you want to do you can run two Port's one with live money and one paper trading only. Use the paper-trade one for timing the live one.  

 Marco,  being able to simulate this using a couple of Sim's would be a fine addition.

Bill

It would be great if a sim’s equity curve could be used as a custom series! Or to enhance the GetSeries() function to accept names of sims / ports in addition to the predefined series.

Analysing the equity curve is the best way I have found for timing when to be in and out of the market. In the simplest of terms - if your curve is going down your system isn’t working so get out and don’t come back until it’s going up again.

Even simple MA crossovers are highly effective and allow me to sleep at night. No, they can’t protect against a flash crash or terrorist action and the like but they saved me in 2008 and other lesser disasters.

At present I use Excel and a program called Market System Analyzer and have done so for a number of years. MSA is really a position sizing program which is useful but doesn’t really fit my needs. Excel would, I’m sure, be great if I was a programmer - but I’m not!

If P123 could program equity curve analysis that would (from my perspective at least) be fantastic. I do appreciate it is a particularly difficult task. Similarly, access to the equity curve as a series could be very useful.

Much simpler would be to enhance the interactive graph by enlarging it and allowing plotting of technical indicators on the equity curve. Ought that not be quite simple to achieve?

Hopefully

Steve

Accessing the port equity curve as a series under the hedging module would be a great addition.

In my case, several of my ports will reduce equity holdings when the markets get stressed. I would like to have a hedge function that reports the cash position as a percentage. That way if a port steps below a set threshold, instead of going to cash, the hedge module could substitute a market short position.

Hope that makes sense!

There is a field called TotalMktVal, the definition of which is “Returns the total value of the port/sim (incl. cash)” If this could just be adjusted keep the history in an array with the most recent being TotalMktVal(0), this could help quite a bit.

–Tom C

That’s like accessing the port’s equity curve w/ the GetSeries() function (if it were capable), right?

Yes, seems like it would be the same thing.

Tom C

It just occured to me that when simulating a port, I don’t know how much of its equity is cash over the sim period. I think I would like access to a port’s equity curve w/ and w/o cash included.

If you want to trade the equity curve natively using P123 you need to vote for feature N. Portfolios to Custom ETF on the top 5 features poll.

http://www.portfolio123.com/vote.jsp?poll=888 (this link doesn’t seem to work - go to POLLS, and look at #4 (number at time of this posting), which was created on 4/10/13)

Otherwise, you will need two ports.

  1. A “tracking port” that generates the tracking equity curve that you download (as others suggested) and use in excel or your trading sw to evaluate trade signals for the model, and
  2. a “trading port” where you actually execute your timing of the curve traded model. As you move in and out of your trading port you will not have a tradeable equity curve, but instead will have the “results” of trading the equity curve, hence you need the tracking port.

With an PortAsETF, you can set up a trading port to trade in and out of the PortAsETF based on equity curve trading, and the model that generates the PortAsETF equity curve is unaffected. The PortAsETF would be a built-in “tracking port” provided by P123.

Unfortunately, feature N is one of the lower ranked requests…

Carl

This is the exported equity curve of one of my R2Gs without market timing but with simple (Donchian) timing applied to the equity curve. The green arrows show where to get in or stay in. The red where to get out or stay out.

Would be good to be able to do this in PF123.

EDIT

Oops, sorry about the size of the image - I forgot to reduce it!


I support this concept and would like to see a simulation be able to time off the equity curve of another simulation.
Alternatively, a stop on the system’s equity curve but a re-entry based on macro, MA cross over or other market indicators, not the simulation equity curve.

Any update on this feature or other equivalent factors to achieve this? Seems like a very useful Market Timing tool.
(I tried the suggested rules in Hedge, but none of them or their derivatives seem to work)

As Marco pointed out, if your portfolio goes to cash, you can’t time a re-entry based on performance, since your performance will be 0 while it’s in cash. Perhaps you could rephrase your question to specify exactly the kind of tool you’re looking for?

Yuval: Walter pointed out a solution for this.

Another solution:

Let us build meta strategy book:

Base Strategy → Signal (= Strategies of today, we do not need to touch them)

Meta Strategy Book → goes long (or short) Base Strategies based on their capital curve and the analysis of the capital curve.

So if the base strategy with going to 100% Cash is the “best” base strategy, the meta book would also go to cash.
But if the Meta strategy book would find a Base Strategy which is not in cash but for example long energy stocks (or short high beta), the meta strategy would
pick this base strategy.

Depending on how you set up the technical analysis of the base strategies the meta book would play different technical setups of the base strategies.

I put this wish in another thread but gave up, so I am very happy that others here want to go to this direction.

What really is striking in this approach.

There are factors that are great for a buy and hold approach (e.g. you simply are long for example “illiquid small cap value momentum” and rebalance regulary).
But there are also factors that trend well. Meaning: If high beta is up for 4 Weeks with with relative strength to the market that means that there is a very high probability that it will go up further (and down if down with relative weakness).
This is not the case for all factors. Investment and Value for example does not trend very well.

Good trending factors:

High Beta
Small minus Big (e.g. Small Caps)
Short Term Reversals
etc.

So what I am doing (not automated) is: I build strategies with factors that trend well and build my own market regime on it.

So automating it would be very interesting.


Stuff like that

Sorry for the rant, but I think to be able to build a meta strategy based on the behaviour of our base strategies and their capital curve behaviour wold be a great add to p123. I also know that this is huge project to automate, so I am fine doing it “by hand” too.

Just wanted to put out that approach and made a video about it: