Stock Market Student

Today I have decided to find a set of highly liquid leveraged ETFs using StockScreen123.   This is quite an easy task as shown below.

The above screen has two rules.  The first rule "ETFMethod = LEVSHORT or ETFMethod = LEVLONG" indicates that I only want leveraged (long or short) ETFs to be selected for my screen.

The second rule "AvgDailyTot(63) > 5000000" is a liquidity rule.  This rule selects only ETFs that have a three month (63 day) average of daily dollar-volume of $5M.  In other words, $5M worth of shares are traded daily on average.

Now by clicking on the RUN button, the screen calculates and displays the results.  46 ETFs meet my criteria.

Now I've decided that there are too many to sort through so I am going to look at specific sectors and choose from there.  This requires a third rule.  The third rule is "ETFSector = xxxxx" as shown below.

 By clicking on the full description you can see all of the sectors. 

As it turns out, only a few of the sectors have ETFs with a high level of liquidity.  From those I tried to select ETFs that have been around the longest (i.e. ProShares over Direxion) and relatively well diversified from each other.  I ended up with the list below as a starting point.

Now I can create a one rule screen that holds these ETFs in equal dollar value as shown below.

Now let's backtest this screen using monthly rebalance.

As can be seen from this graph there was a 45% loss over two years.  This amounts to a profit when short selling the ETFs.  Next post I'll do some further evaluation of this screen.

Take care,

Happy trading

Before going any further with my system development I decided that I need to find an internet-based tool to assist in analysis instead of attempting to do everything using EXCEL.  I believe that StockScreen123 will do the job.  It is *** FREE *** for now while it is in beta.  It is a sister site to Portfolio123, which is very excellent for stock investment portfolios.

StockScreen123 is slightly less complicated to use than Portfolio123 and has the ETF backtest features that I need.  Here is a small demonstration of this web-site's capability.  Lets say I wish to short both the long and short leveraged SP500 ETFs.  Well I can't actually short the ETFs using StockScreen123 but I can demonstrate what happens over time as they are held long.  Here is how it is done.

By entering one simple rule Ticker("SDS,SSO") it tells the screener that I wish to hold equal $ value of the ticker symbols SDS (Ultrashort SP500) and SSO (Ultralong SP500).  Now by hitting the backtest button the backtest page comes up, looking something like this.

It takes a maximum of five steps to setup and run the backtest.  Some of these parameters may be suitable in their default state and do not need changing.

1. When I am in development mode I always set the slippage to zero.  Once I have a serious system I will set the slippage to something realistic.
2. I set the price to Next Open.   This means that the backtest will use the next opening price in it's calculations.  An alternative is to use Next Close.  This means the backtest will use the closing price for the next day as opposed to opening price.  This can be a better choice as there is likely less slippage than opening price.
3. Set the date range.  In this example I am doing a 2 year backtest.
4. Set the rebalance frequency to 4 weeks.  This means that every 4 weeks the holdings will be adjusted so that each ETF has equal $ value.
5. Click on the Backtest button.

The backtester then returns with a graph of performance over time and month-by-month summary in table form.

Note that there is a steady decline in return over time (except for the fall of 2008).  This is about as I would expect as leveraged ETFs tend to lose value over time.

As shown above the table summary gives a month-by-month accounting of the screen performance and an overall summary.

 Next post I am going to try backtesting 5 3x leveraged ETFs (long and inverse) to see if I can smooth out the decay.

All for now,

Steve