Introduction to Meta-Strategy

In Portfolio Boss, you can treat each strategy as if it were an Exchange Traded Fund (ETF). That means you can buy or sell an entire strategy as if it were a normal financial instrument.

Just like an ETF, each strategy bought/sold contains a collection of instruments within it. This is the purpose of a Meta-Strategy. So a Meta-Strategy is just like any normal strategies, but instead of individual stocks or bonds in its Portfolio, it contains a list of strategies to be bought or sold.

If a Meta-Strategy enters a Position, it's buying all positions of a Strategy. As long as that Position remains open, Portfolio Boss will give you Buy and Sell signals based on that strategy; in essence you're following that Strategy. Think of it like you're holding an ETF position, but you also buy and sell the individual stocks as well.

If the Meta-Strategy sells a Position, it means you're selling all Positions of that strategy, and the released fund will either be parked on the Meta-Strategy's Cash Equivalent (you'll buy the Meta-Strategy's Cash Equivalent), or be used to enter another strategy.

Note that if some strategies contain the same instrument, PB will not group them into one trade, just like when you're buying two separate ETFs. So it's preferable that each strategy contains different portfolio from the other strategy. For example, one strategy is about Technology Sector, while the other is about Banking Sector.

Meta-Strategy can only be created with certain licenses. If you wish to upgrade your license, please contact our Customer Support team at support@portfolioboss.com.

Creating a Meta-Strategy from Scratch

To create a Meta-Strategy, go to the “Strategy Management” page, and click the “New” button on the upper left corner. A dropdown will appear, and select “Strategy” from here.

On the dialog box that appears, you can give a name to the Meta-Strategy, and select which group it will belong. Most importantly in this dialogue box, the “Strategy Type” parameter must be set to “Meta-Strategy”. 

Now, press OK and you'll be taken to the “Backtest Strategy” page where you can customize the Meta-Strategy's parameters.

Customizing the Meta-Strategy

The “Backtest Strategy” page is perfectly identical to any normal strategies, in which you can add Buy Filters, Ranking Rules, Sell Filters, and any System Settings for this strategy.

The only difference lies in the Instruments Panel. As explained before, instead of adding a Portfolio of financial instruments, you add a list of strategies here. To do so, click the “Add Strategy” button on the top-right corner of this panel.

A dialogue box appears allowing you to select which strategies to buy/sell. Keep in mind, you can only select basic strategies here (either bundled with your license, or the ones you created yourself), not other multi or meta-strategies. 

You can search specific strategies by typing a name on the “Search strategies…” above. Once these strategies are selected (their checkboxes enabled), press the OK button and these strategies are listed on the Instruments Panel.

You can edit any of the listed strategies by clicking their name (highlighted blue). 

A dialog box will appear allowing you to edit the strategy's parameters. Keep in mind, editing a strategy through this dialogue box also changes the actual strategy.

Now, the most important thing about Meta-Strategy is this: as you already know, the Buy, Rank, and Sell filters normally look at an instrument's price history. But when it comes to a Meta-Strategy, these filters look at each strategy's gain (which you can see on the Chart Tab if the strategy is backtested independently). 

So for example, if a strategy is experiencing a drawdown i.e. its equity curve falls below the 200 days SMA Sell Filter, the Meta-Strategy will exit that strategy's entire positions. This way, think of the meta-strategy as a layer above your basic strategies, to improve upon them. Since they're analyzed as financial instruments, you can decide when to trade them, so you can have an even smoother ride.

Backtesting the Meta-Strategy

Now that the Meta-Strategy's various parameters are set, you may backtest it as usual by clicking the “Start” button.

A confirmation dialog appears, allowing you to write a memo describing this backtest. If you don't want this dialog for future standard (manual) backtests, tick the checkbox “Don't show again”. To actually start the backtest, press “Continue”. The memo you wrote can be seen under the Manual Results Tab (once the backtest is finished):

The backtest may take quite a while since we're dealing with many strategies at once. Once finished, the resulting backtest reports are perfectly identical to a normal strategy, but instead of actual instruments, you'll see the strategies listed on the reports. 

An exception is the Cash Equivalent for the Meta-Strategy, which is listed as a normal instrument.

You can click a strategy's name (from the Positions Panel, for example) and the Price Chart for that strategy will show up, showing the candlestick chart as if it were a normal instrument. The candlestick bars represents the percentage gain of that strategy; you can also see the various indicators overlaid in this chart. 

The only thing missing is the Volume Chart data, since this is a strategy, not an actual ETF which actually gets traded on the Exchanges.

There is one particular panel in this Meta-Strategy that you must know: the Trade Signals Panel. As said before, instead of instruments, a Meta-Strategy buys/sells strategies. But the trades that you must do, actually involve buying/selling actual instruments. 

This is reflected by the various buy/sell signals on this panel. As you can see, it lists all Positions within the strategies that have been selected for a trade. It shows which Positions belong to what Strategies.

Note:

If your license includes the Divine Engine capability, you may be able to use the Random Search Algorithm on your meta-strategy. The Divine Engine takes it like any basic-strategy, except its instruments (portfolios) are its component-strategies. If these component-strategies were also created through the Divine Engine, you're binding these high-performing strategies (“instruments”) into a greater strategy, which is also optimized by the artificial intelligence. Hence it's a multi-layered winning system!

Pay attention to the property “Rule Set Parameterization” (on the Backtest Parameterization Panel, once the Divine Engine is enabled):

• The first option, “Random selection”, allows the Divine Engine to slap random new filters (with random parameter-values) to the strategies it creates. Existing rules & filters will be “fine-tuned” to find the best values for them.
• The second option, “Manual”, simply fine-tunes the existing rules & filters (based on your parameterization) to find the best values for them.

Please contact our customer support if you wish to upgrade your license.

Back to Top