I think the financial equivalent of jumping shark is Wall Street creating an ETF based on your theme.

If so, cybersecurity has arrived.

The ISE Cyber Security Index provides a benchmark for investors interested in tracking companies actively involved in providing technology and services that are designed to protect data, networks, hardware, software, and other cyber-dependent devices from unauthorized access and attacks. The index includes twenty-nine constituent companies, including VASCO Data Security International Inc. (ticker: VDSI), Palo Alto Networks Inc. (ticker: PANW), Symantc Corp. (ticker: SYMC), Juniper Networks Inc. (ticker: JNPR), FireEye Inc. (ticker: FEYE), and Splunk Inc. (ticker: SPLK).

Before you invest your life savings in ETFs, listen to Vanguard founder Jack Bogle: “The ETF is like the famous Purdy shotgun that’s made over in England. It’s great for big game hunting, and it’s great for suicide.”

Two interesting things to look at in ETFs are fees and weighting. The fees on this puppy look to be 0.75% – outlandishly high. For comparison Vanguard’s Dividend Growth ETF has a 0.1% fee. It is true that with foreign ETFs the fees are higher (to access foreign markets), but I do not know why HACK should have such a high fee – the shares they list are liquid and widely traded. Foreign issues themselves do not seem to dictate such a lavish expense ratio.

As of October 30, 2014, the Underlying Index had 30 constituents, 6 of which were foreign companies, and the three largest stocks and their weightings in the Underlying Index were VASCO Data Security International, Inc. (8.57%), Imperva, Inc. (6.08%), and Palo Alto Networks, Inc. (5.49%).

I cannot tell how it is weighted but if they follow the weighting on ISE then investors will wind up almost 10% into Vasco. The largest members of the index, per ISE, are:

Vasco: 9.17%
Imperva: 7.57%
Qualys: 5.48%
Palo Alto: 5.35%
Splunk: 5.18%
Infoblox: 5.04%

That is near 40% in the top six holdings – pretty concentrated. The old school way to index is to weight by market capitalization, but that has been shown to be imperfect because size alone does not determine quality. The preferred weighting for the last few years (since Rob Arnott’s work) has been by value, which bases the percentage of each holding on value metrics like P/E. There is considerable evidence that this works much better than market cap. But we still have a problem: many tech companies, especially new ones, have no earnings! From reverse engineering the index membership it looks like they are using Price/Sales for weighting. For example:

Vasco has a Price/Sales ratio 6.1.
Palo Alto has a P/S ratio of 13.5.

Vasco has about twice the weighting of Palo Alto because it is about twice as cheap on a Price to Sales basis. This is probably not best way to do it, but it is probably the best available way because market cap is flawed and would miss all the upstarts. Due to lack of earnings value metrics are a non-starter. The weightings appear roughly right per Price/Sales, but I could not get the numbers to work precisely. It is possible they are using an additional weighting factor like Relative Strength.

Needless to say, this is all in the spirit of “As the Infosec Industry Turns…” and not financial advice of any kind. This is not a recommendation to buy, sell, or hold any of the issues mentioned.

In the meantime remember the fees, and this from Jack Bogle: “Performance comes and goes but cost goes on forever.”

HACK SEC filing