Market Data Questions
Rydex Asset Numbers
I have a question about using the asset levels of Rydex funds. There are so many new ETFs that have recently come on the market that I wonder if using the Rydex asset levels with absolute dollar value is still useful. Rydex has however taken in about $2B in assets over the last 3 months, so perhaps, the indices remain useful. I now use the ETFs though to go short.
We were among the earliest adopters of using Rydex fund asset levels as a sentiment indicator. The first person we are aware of to start doing this was Steve Todd, of the Todd Market Forecast (see also our article on "The Steve Todd Test").
Rydex was the first fund family to introduce a bearish fund, the Ursa Fund, which moved inversely to the movements of the SP500 Index. They were also unique in their practice of allowing traders to move into and out of their funds on a daily basis, with no withdrawal fee or trading restrictions like those seen at other fund families. The third interesting innovation was that Rydex published the total assets invested in each of its funds on a daily basis, by a recorded telephone message.
Steve Todd quickly realized that this data had importance as a sentiment indicator. By comparing the level of assets invested in the bullish Nova Fund versus in the bearish Ursa Fund, one could gauge investor sentiment about the market. These indicators worked great for many years, but the changing dynamics of both the mutual fund industry and the financial markets in general have diminished that utility.
One of the big changes has been the explosion in the number of exchange traded funds (ETFs) which offer the index tracking and unlimited trading utilities that were once unique to the Rydex funds. In addition to the ETFs proliferating, Rydex has also seen competition from the likes of ProFunds with its own set of bearish funds, which serves to draw money away that might have gone to Rydex. But the flip side of this is the concept that increasing competition actually can be shown to increase the total size of a market, such as when a Burger King goes in across from a McDonalds. Not only does that McDonalds restaurant not see its lunch business cut in half, but the competition can actually increase their business by attracting more people to the area. It is a paradoxical conclusion; a lot of econ PhD papers have been written about this. Thus, the mainstreaming of these seemingly exotic funds can bring more money into Rydex which might never have gone there.
Trying to model all of those changes can muddle the brain, and the best we can do is acknowledge their impact, and watch how the indicators behave. The variability does make it inappropriate to assign absolute levels to these funds' assets for where "high" and "low" are, but we can monitor the current asset level relative to the past and derive informational insights about the market. I'd never trade solely off such information.
I can say that I have noticed Nova's asset level does not seem to work the same, and nor do the dynamic (2 beta) funds' assets. But Ursa's raw asset level still offers some utility as a sentiment indicator, as does the Bond Fund, which is why we feature them when they have something interesting to say. The asset levels of the Rydex sector funds are also still useful, especially when examined in comparison to the total amount of money invested in all Rydex sector funds, so that one can gauge which sectors investors think are the hot ones, and which sectors are out of favor.
As an aside, I met with reps from PowerShares at the January 2007 MTA conference. They are bringing out 2-beta (200% leverage) positive and inverse ETFs on the Russell style indices, e.g. Russell 2000 Growth. They are also bringing out a set of 2-beta sector ETFs. I asked them if I could get access to daily data on the total float for their ETFs, since that would presumably work similarly to how we use the Rydex asset data. For reasons of competition, they keep that information close to the vest.