I have tried simulating daily rebalancing and have developed some great trading systems. However I am unable to use daily rebalancing in Portfolio generation. What is the point in being able to simulate without being able to make a portfolio out of it?
Maybe I am missing the point here or doing something wrong?
My opnion to Steve’s (Stittsville123’s) post is that you have to start with how you manage a portfolio. This is the reverse of where I was - develop a good sim, create a portfolio and trade your brokerage account to the recommendations. Optimizing sims that you cannot or will not follow in the real world is not the answer.
I run multiple ports with weekly rebalance and use stop losses. There are some other price related sell factors but most are stops. I generally react to a stop in my brokerage account by mostly holding cash until the new rankings are available on the next weekend. I do not want to use “old” ranks to buy a replacement position. If the stop hits on Monday I may feel differently. If it hits late in the week, I wait. I almost always wait. I then rebalance the port on the weekend with fresh rankings and most often find the stopped stock is a sell recommendation. I accept the sell, edit the price to reflect the stop price (generally higher than the sell recommendation) and buy a recommended replacement.
Am I really managing my ports as daily or weekly rebalancing? While I first thought I was closer to daily, I now think I am closer to weekly. To manage a port to match a daily rebalance sim, I would have to check the port each day and buy a recommended replacement based on the previous weekend’s rankings. Then buy and edit the port manually. This is OK but I do not think I will actually do that very often; not because of the effort but because of the “old” rankings. Maybe weekly rebalance is a better simulation of the actual management of a brokerage account. The simulated returns are generally better but the proof is in the actual portfolio, not the simulation.
While I am at it the new Transaction Pricing (Pervious Close, Next Open and Next Day High/Low) raise the same issues. There has been a lot of study of the differences between previous and next in the past with the consensus being it did not make a significant difference. I suspect that most felt that since they were rebalancing on Monday, Friday’s price was stale. But what do we really do in our portfolios? I tend to actually buy and sell on Monday morning when I can look at what is happening to a stock. I almost always follow the system but I like to see if there is something strange happening. This means that I probably execute my actual trades closer to the average than the Monday open.
Given all of the above, right now I am a weekly rebalance and a Next Day High/Low guy for sim development. I think this best matches the way I actually run my real ports.
I believe they are still working on daily rebalancing for portfolios. In the meantime, you can do daily rebalancing manually. Just go into your portfolio, click on Rebalance, then click on Get Recommendations. If it comes up with a recommendation, you have the option of accepting it or not.
thought i got the impression somewhere daily rebalancing produce as good if not better results. based on a little testing that didn’t seems to be wrong.
My experience is that daily rebalance generally reduces the performance of sims. Check “Effects of new Sim tools”
(http://portfolio123.com/mvnforum/viewthread?thread=842) for a discusson on the new sim tools. My results are similar to those of others.
I wanted to first do the daily update of fundamentals. At the moment ranks only change once a week. I think the daily rebal would make more sense if the fundamental ratios change every day. The daily fundamentals project is starting soon.
Yes, I see that makes sense and it would be a good addition ( Ithink, without testing it). I do have a Port that I have just started that looks at price daily - and does better than looking at weekly. It would just be nice to get an email each day telling me if I have to worry about doing anything - but of course not absolutely necessary. I’ll just have to be patient.