We all know there are different phases to a market :
Bottoming , bull , topping , bear just to use the simplest categorization
It is always easy to spot where we are . Right now we are either in a bull or a topping phase .
What has this to do with portfolio123 and our methodical approach here .
The question I have would be :
should we use the same buy and sell rules for every phase ? We at this stage apply all rules consistently . My belief is that we will be able to improve results should we be able to be more specific as to which rules to apply and when eg one would definitely use different buy rules at the bottom of 2009 than now . Maybe high beta at bottom vs low beta now ?
I believe this enhancement would be easy to implement in portfolio123 : conditional buy and sell rules
I hope I will get some support for my idea from the community .
Hugh,
thanks for your comments. They sound very good.
But I am afraid that the “phase” of the market can only be determined in hindsight. The market is a chaotic beast and “phase” (or “timing”) have been elusive for the most part. As you know, there is always plenty of disagreement and opinions as to what “phase” we are in. And nobody to my knowledge has ever shown that this determination can be made reliably (or even approximately).
Unfortunately.
Wern
I have tried this also with several ‘mkt state’ switching systems for the past 5 years now…using both ETF’s and market regime node switching ranks, and evolved conditional rules, etc.
I’ve found that it’s much easier to accomplish this objective with a ‘book’ than with single systems. We can combine focused, single systems that look to invest when certain conditions are met. The ability to define our own series lets us do this very well (once it’s live). Nesting systems with very different timing rules and / or conditionals that turn ‘on’ and ‘off’ lets us do this.
For example…when small caps are outperforming large caps, we are more heavily invested in small caps because an additional system ‘kicks in’.
The major issue right now in terms of managing all this with P123 is the lack of a more refined ‘master’ control ‘book’ system that allows us to use rules to alter the total ‘book’ based on underlying subsystem performance. I would like to be able to alter subsystem weights based on trailing period performance, correlations, volatility, etc. I would like to be able to alter ‘leverage’ amounts at total book level as well. I have posted feature requests on ‘more sophisticated book’ tools in the past. But…this is (for me) the better way to achieve this end goal of dynamic market allocations based on ‘market state.’
In total agreement with all you say . I started with a request at portfolio level because I think it would be an easy one to implement for the development team but your more elaborate approach should be the end goal .
I would love to be able to vary the weights of different systems, individual stocks and ETFs dynamically based on some conditional systems.
I have one system that works best in the late stages of a bull market. It’s doing wonderfully right now. It does not do as well in the early stages. I have another system that does fantastic in early stages but okay in late stages. Weighting systems based on momentum would help quite a bit in timing when and how much to invest in each system.
By varying the weights of individual ETFs we can almost eliminate volatility. The idea is to vary the weights of the individual ETFs based on expected return, risk and correlation using current market data as inputs.
The way I have approached this, and I am currently trading it, is I set up 5 Sims with restrictive buy rules and low correlation. Each Sim gradually stop buying when its specific area of the market declines, gradually going to near 100% cash. There is little overlap between the periods that each Sim is going toward or is in cash except for the deeper periods of the 2 recessions. I set up a Book Sim to test various combinations of low correlation Sims.
I then created Ports from the Sims that showed the smoothest curve in the Book Sim. In trading the 5 Ports, the cash that is freed up from one Port is used to buy stocks from another Port as it starts buying back into the market. I manually rebalance the Ports and sell/buy the recommendations from the combination in a single account Portfolio. I don’t bother with rebalancing a live book, but occasionally rerun the Book to see how it performs out of sample. Since I started this approach, nine months ago, I have been above 90% invested even though a few of the Ports have been as low as 30% invested.
Seems to me that at the begining of a bull market, after a severe downturn, value stocks do the best, because they bounce back the most from being undervalued. Investors start to realize that the sky has not fallen and these valuations are too pessimistic and adjust accordingly. This is where Warren Buffett is a kid in a candy store.
Towards the end of a bull market, value stocks do poorly because they are out of favor, and growth/momentum stocks do the best. At this point, stocks are overvalued but investors don’t care about valuation and keep pushing stocks higher. This is where Warren Buffett sits around and tiddles his thumbs.
As I have asked in the past, it would be great if there was an option within Book that will take cash from ports that are out of the market (partially, or fully out of the market) and redistribute this cash into the remaining active ports in the book. This would allow you to be 100% invested. So, cash from the out of favor ports move to ports that are in favor.
