Portfolio Heat

Discussions about the testing and simulation of mechanical trading systems using historical data and other methods. Trading Blox Customers should post Trading Blox specific questions in the Customer Support forum.
Post Reply
Carlopp
Contributor
Contributor
Posts: 3
Joined: Sun Mar 12, 2017 9:05 am

Portfolio Heat

Post by Carlopp »

hi I'm new too trading blox and wanted to know how i would use Portfolio Heat in the trading blox Risk Manager block meaning i just want to be say e.g. 25% risk/heat of total equity at the one time its a protection mechanism from price shocks .. Van Tharp and Ed Seykota talk about this not sure how to explain it.
Tim Arnold
Site Admin
Site Admin
Posts: 9015
Joined: Tue Apr 06, 2004 1:41 pm
Location: Boston, MA
Contact:

Re: Portfolio Heat

Post by Tim Arnold »

Yes, that's a great idea. Couple of things to consider about 'heat' or risk. When we add up the risk of a portfolio, we loop over each instrument and compute the price difference between the close and the current stop price. So be sure your system has stops, and places those stops into the market every day.

The first thing you might do is block new orders if the risk /equity percent is above some threshold, say 25%. In this case no new orders will be placed but that does not stop the risk from going up or down, above or below your threshold. Each day as the market moves, and the stops are adjusted, the risk is recalculated.

Another consideration is placing many stop or limit entry orders on the same day. You don't know if they will be filled, so when placing the orders you can't keep under a minimum risk threshold. So when the current risk / equity is less than a certain threshold, decisions need to be made about new orders. When using market on open orders this is an easy problem, in that you can compute the risk for each order knowing it will be filled, and rank the orders by strength or other criteria only placing the 'best' orders and keeping under the risk threshold. The attached block has a 'Process Fills' option which is nice for back testing -- checking the risk as each order is filled, but needs to be done manually in live trading.

At the end of a day, you could take a look at your risk and decide it's too high, over a threshold percent of risk / equity. In this case you could reduce your stops to make the risk of each instrument less, or you could reduce the position size of each position.

Many of these implementation decisions will depend on your system design, how the stops are computed, how the orders size is computed (fixed fractional), types of orders, and desired behavior. Even if you adjust the risk one day, the next day will be different and the risk will be higher or lower -- so the risk will never be capped at a particular level or a flat line.

For starters, attached is a Portfolio Risk Manager block that deals with a number of the above issues.

Max Risk Set to 30%:
totalriskprofile.jpg
Attachments
Portfolio Risk Manager.tbx
(6.46 KiB) Downloaded 294 times
Carlopp
Contributor
Contributor
Posts: 3
Joined: Sun Mar 12, 2017 9:05 am

Re: Portfolio Heat

Post by Carlopp »

Thanks Tim for the information very helpful.

Also i contacted premium data and they asked me, If Trading Blox can’t do this but it can work with ASCII/CSV data instead, then you can use our Premium Data Converter (“PDC”) program and set the Periodicity to weekly.
Chuck B
Roundtable Knight
Roundtable Knight
Posts: 481
Joined: Thu Apr 17, 2003 6:34 am

Re: Portfolio Heat

Post by Chuck B »

One other (of the many) aspect of portfolio heat is the consideration of scaling out (reducing open risk), using individual asset risk percentage of the total portfolio, during the progress of the current bar. For example, consider that you can calculate tomorrow's (next bar's) exit (stop) location for any current bar in progress. Hence if there is a huge move in one asset (or in a group of reasonably correlated assets), you can continuously calculate the potential stop for the next bar. This is exactly true in the case where the exit is based off the extreme of the current bar (i.e. such as a trailing exit of some sort that is a function of the high (low) of the current bar). In this case, you absolutely know the location of the exit for the next bar for every value of the current bar in progress.

Hence, you can consider scaling down risk as the asset risk becomes greater than some fixed (variable perhaps too) value of total portfolio risk. If you have say 20 assets in the portfolio, and one of them moves so far so fast in the current bar that it's risk exceeds this value, you automatically start exiting some of that asset until it's risk level to the portfolio is below your threshold. It's an "exercise for the student" to determine what levels of risk are desired in this case of course. If you perform this function in real-time, it does allow you to scale out, perhaps completely, during "wicking", blow-off type price moves which are otherwise ignored with a static system.
Post Reply