How to Start an ETF

I gets lots of questions over email and in person on a handful of topics.  Usually I try to incorporate all of them into a longer written piece that includes my personal opinions and experiences that readers can then reference.  You can find some older ones here on the following topics:

Global Tactical Asset Allocation & Asset Allocation

Self-Publishing a Book

Income Investing, Risk ParityBubblesPensions, Black Swans, and Valuing Global Stock Markets.

One of the themes I get a lot of questions on is the process of starting and managing an ETF.  We have somewhat of a unique perspective as a user, investor, sub-advisor, and advisor of ETFs.  I thought I would summarize all of the questions I get on a regular basis – if you have any more questions fire them over and I’ll incorporate into the article.  But first, some background.

Separate Accounts

We started our firm in 2006 with lots of ideas, but no definite roadmap as to what structure would best suit our firm.  We began with separate accounts.  Separate accounts have numerous benefits, namely a) a relationship with the client, and b) ability to tailor the strategy to their particular risk tolerance or tax situation etc.  Those both can also be drawbacks of course, as client relationships can be time consuming and headache inducing.  However, they can also benefit the client as they have someone looking over their investments and potentially talking them off a ledge and doing silly things at the wrong times.  Raising assets through separate accounts is driven largely by relationships and marketing.  For an example, take a look at Fisher Investments, one of the all time great marketing companies via direct mail (and now in print and online).

We still do separate accounts but this is largely for high net worth investors where we can offer some personal touch.  Look for a big announcement here soon.

Hedge Funds

We launched two hedge funds back in 2009, largely for clients looking for more aggressive strategies that included leverage.  We closed the first one when we launched our first ETF, but still have two open, including an insurance dedicated fund for a family office.


If an ETF launches with no assets and fails to raise any, you are probably on the hook for ~$150K in expenses per annum.  That is a lot of checks to write. Launching an ETF is very much like a rocket getting out of gravitational orbit, it needs an escape velocity of perhaps $30-$50 million to get profitability.  There also becomes a chicken and egg problem – ETFs with low assets and volume are often seen as orphans as larger investors have AUM and volume minimums, but the ETF cannot get larger without assets and volume.  This is why you have seen in the early 2000s about 80% of ETFs launch successfully, while in recent years that number is closer to 20%.  Unless you KNOW that an ETF can be seeded with, or raise enough assets to get to this $30-50M range quickly, it has a much higher chance of failure, especially now that there are so many products.  Some firms adopt the shotgun/VC approach, hoping a small number of very large ETFs will subsidize all of the small ones.  Other firms adopt a rifle approach with only their best ideas.  Both can succeed or fail.

Index Provider

We first considered launching an ETF with Claymore (now Guggenheim) when Forbes was looking to move into the ETF space.  Our launch date for “IVY” was 12/2008, talk about bad timing!  However this would constitute the first method of launching an ETF – as index provider.  Typically this is the least effort, you simply provide an index that updates/rebalances, but also the least rewarding monetarily.  Typically the index provider can earn as little as a few basis points on the low end, to perhaps 30 on the high end.  Lots of ETF companies will license an index if they think there is product demand.  The PowerShares Research Affiliates relationship is a good example of a very successful advisor/outside index provider.  Lots of companies will track indexes for a fee of $2,000 -$30,000 like Solactive, S&P, MSCI, NASDAQ, etc.

Sub Advisor

Subavisory is usually when a RIA that does not have, or want the SEC exemption to launch ETFs, partners with a company that does. The model is very similar to the publishing industry where a content provider parters with a firm to provide infrastructure support.  There are a number of players here including AdvisorSharesExchange Traded ConceptsAlpsETF Issuer Solutions, and Canvas.  If you are a manager looking to partner with one of these firms, they will send you a basic proforma that will ballpark most of the costs and profits as you raise (or don’t) assets.

The economics are more favorable than index licensing (generally), but may also involve upfront fees ($0-100k) or exposure to monetary risk if the ETF does not raise assets.  Typically the subadvisor receives somewhere in the ballpark of 50-90% of the profits from the ETF, depending on other costs and risks taken.  

Other pros of subadvisory is not having to build the infrastructure that goes along with launching an ETF and the subsequent costs like board meetings, hiring a CCO, and other duties.  

Cons of subavsiory include loss of control over the product, revenue sharing, branding confusion, etc.

This option is usually a great choice for those who are looking to only launch one or two ETFs, and not have to deal with pursuing their own exemption.

Advisor/Sponsor (also more economics) –

You can find a full list of advisors here:

You can find the list of advisors by AUM here:

Becoming a sponsor is time consuming and costly.  Expect 12-16 months, and (mostly legal) expenditures of $200-300,000.  However, once you have the exemption you can take ETFs to market in approximately three months if the strategy is plain vanilla, or six to infinity months if not.  I expect the SEC to streamline this process in the future.

Pros include autonomy, full revenue capture, and total control over the product.  

Cons include costs, infrastructure buildout, and time to market.


This is a new business model where real money instiutions have an ETF idea, and agree to seed it to see it come to market.  Some recent examples include Arizona and Fisher.

I am surprised more taxable real money institutions (like family offices) don’t launch their own ETFs for their internal, active strategies for two reasons – tax efficiency, and potentially revenue capture.  

When will we see the first endowment or family office ETF?