Exited Alternative Asset Opportunities

After buying Alternative Asset Opportunities (TLI) almost exactly one year ago I exited my position today at a very small loss (-2.4%). Fundamentally nothing really changed between today and one year ago. What changed was my understanding of the traded life interests. If you start with my oldest post on TLI and read the follow-up posts chronologically you see how with every update I realize that my model of the expected future cash flows needs to be refined, and obviously always in the ‘wrong’ direction. After my latest post on the fund I realized – with the help of a reader, thanks! – that the model was still too optimistic. With my perceived margin of safety evaporating the logical thing to do was selling my position.

I always write that readers should do their own research, and if TLI doesn’t prove the point I don’t know what will. I have compiled a quick list of all the things I missed initially:

  • Update, Nov ’12: the average age of the portfolio doesn’t increase with one year/year.
  • Update, Nov ’12: the ratio of females and males doesn’t remain constant
  • Update, Oct ’13: added assumption that premiums increase 5%/year
  • Update, Oct ’13: realization that some policies lapse when holder reaches age of 100
  • Update, today: 5% premium increase per year too optimistic, probably 9~10%/year
  • Update, today: reducing average age with one year doesn’t increase LE’s with one year

So keep this in mind the next time you read a post on my blog! Especially the last point in the list matters for the valuation. While quite obvious in hind sight I didn’t realize how big the difference was. If you assume that insured group is as healthy as a ‘normal’ 3 year younger group you only add 1 year to their LE. Given that the new LE provider is expecting that the insured group could have a LE that is two years above the previous estimates it’s now easy to see a scenario where you actually start losing money (and given the historical profile of policy maturities it certainly looks like they could be right). TLI simply doesn’t appear to be the opportunity I once thought it was.


No position in TLI.L anymore

9 thoughts on “Exited Alternative Asset Opportunities

  1. Tom

    a) Surely all of the above is baked into TLI’s model?
    b) Even if TLI’s model is over optimistic, it’s at a 20% to NAV, which uses a discount factor of 12%. How bad can it get?
    c) They are pretty much net cash generative and in a position to buy back shares, which they have said they will do. This will be accretive.

    If you think of this as a fixed income like asset, offering some insurance against equity markets falling, I don’t think this can be dismissed.

    1. Alpha Vulture Post author

      a) Most of it is, except that the LE’s are probably too optimistic. If you would add two years to the LE you have multiple factors that significantly reduce value: 1. you pay premiums for a longer period of time 2. you pay high premiums because they increase every year 3. you get your cash flows further away in the future and 4. you pay management fees for more years.
      b) The NAV doesn’t account for management fees; not all cash flows go to shareholders. $1 million/year in fees does count for a fund this small
      c) This is only true if shares are undervalued. That might not be the case…

  2. Packer

    I think you have left the Titantic before it sinks. These guys are using the standard US life tables and just now are using a different LE provider. The rest of the industry has been using other provides for years and the best investors (the ones who make money on these things versus reported losses) use there own internal data bases that have LE longer than those provided by LE providers. To give you an idea of the types of discount rate that equates external LE providers to some of these more conservative internal models, the provided LEs have to have discount rates in the upper teen low twenties to equate to a internally developed data base in the 12% range. The equivalent here would be a 20% discount rate on the new LE adjusted cash flows.

    The adverse selection bias in these portfolios can be incredible and the impact of changing LE assumptions a few years can be very high due to the fact that no only is you payoff stretched out a few years but you have to put more cash in. The adverse selection comes from the fact that a good number of the policies are from relatively wealthy folks who have used the for estate planning purposes. These folks have access to great health care and typical keep in shape.


  3. Packer

    I tried about a year ago on the Corner of Berkshire and Fairfax board – http://www.cornerofberkshireandfairfax.ca/forum/general-discussion/what-would-you-guys-buy-today-givei-100/msg94943/#msg94943

    This is a very difficult asset class and looks alot easier than it is, sort of like buying LEAPs on undervalued stocks. The main issue I think is the mark to model is very complex and has numerous unknowable assumptions in order to develop a value and it is hard for an outsider to assess the reasonableness of these assumptions because the risks you are dealing with are non-linear and the benchmarking data is not available to anyone but insiders (those who have understanding of the health and genetic history of the insured and pools of this type of data to make an assessment). Even if you have this information it is difficult because LE are constantly getting longer with improving technology.


    1. Alpha Vulture Post author

      Guess I can only blame myself then :p. I thought there would be a sufficient margin of safety for those issues, but my model was just too flawed… this one needed some time before I was able to appreciate the complexity of the asset class and how big the negative impact could be if LE’s would be underestimated by a couple of years.

  4. Packer

    You are way ahead of most investors as you got out with a small loss. In my first set of investing mistakes, they were complete wipes outs. I literally went down with the ship to $0 as I didn’t see/believe the mistake I had made and in some cases compounded the error by averaging down.


    1. Alpha Vulture Post author

      Thanks for the encouraging comment, but I also think that the past couple of years have simply been a kind environment to mistakes: hard to go really wrong if everything goes up…


Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.