Monday, 7 March 2011

Correlated risks revisited

After the September quake, Peter Cresswell pointed to some worries about the New Zealand Earthquake Commission (EQC)'s balance sheet.* Specifically, it's heavily invested in New Zealand domestic assets, mostly government securities. I checked the 2008/2009 annual report - about two thirds of its asset base consisted of government securities. This seemed a pretty odd way of setting things up. A big earthquake hits, and folks are going to start demanding a risk premium on New Zealand government securities. That means their value drops at the same time as EQC has to offload a pile of them to pay out claims. I'm not a finance guy. Maybe there's some high tech fancy finance reason why this is a fantastically good idea. But it seemed then and still seems remarkably silly.

I've now checked the 2009/2010 annual report, which was signed off before the September quake but not published 'till after I'd posted in September. As of then, just over 69% of the asset base was in government securities.

The dollar dropped more than a cent right after the quake. If EQC had been heavily invested in foreign securities, that would have been a good time to ditch some of those assets and buy cheap New Zealand dollars. Instead, it's going to have to sell domestic assets in a buyer's market.

I hope that EQC adjusted its portfolio subsequent to the September quake.

Seamus had proposed a tidy solution: if politics means that EQC has to ignore financial sensibility and invest domestically, then at least it could invest in earthquake countercyclical assets. Construction companies would presumably fare well, relatively speaking, after a natural disaster. That recommendation probably needs some work - Fletcher Building is up overall, but not particularly in response to the February earthquake. Total market capitalization of the building sector is down since the September quake.

Maybe one of our future Finance honours students can mine through NZX data to build portfolios of earthquake countercyclical stocks. We now have two decent events.

First best remains having the whole thing invested overseas.

* For overseas readers, here's the two minute summary of the EQC. If you buy property insurance, as a homeowner, you pay a small levy on top of your normal premium for natural disaster insurance. The Earthquake Commission - a government agency - then takes on the first $100,000 of property and the first $20,000 of contents damage in the case of a natural disaster. The premium is not risk adjusted: $0.05 per $100 in value insured to a maximum premium of $67.50 for a maximum of $100K in property and $20K in contents insurance. We'll see what's left in the kitty after these two events.


  1. I guess we could also complain about the incentives resulting from their failure to risk-adjust the premium.

  2. Could it be that they are holding the government securities to allow a redemption at face value should a crisis hit?

  3. Your right about the risk on risk element, but I suspect that the reason EQC holds NZ govt securities is to avoid forex risk, but that doesn't seem to be enough of a reason... If EQC held offshore assets and it was called to payout the NZD would have fallen, as it did after the Chch earthquake... which would mean that it would get a better return on its foreign holdings - all other things being equal...

  4. @kimble if by that you mean redemption by the govt, that makes things worse, not better.

    @Horace Dump NZ assets when the dollar seems strong, buy foreign. Perhaps some currency risk but they could always buy options if seriously worried about it.

  5. Never said it was good, but I think it would explain the weight in the portfolio.

  6. The holdings are determined by ministerial direction see:

    Specifically see section 9:

    9. The Commission must consult with the Minister if it intends to modify the portfolio composition from the following:
    - NZ Government securities;
    - up to a maximum of 35% of the market value of the Fund in global equities; and
    - up to a maximum of $250 million of New Zealand bank bills.

    So no more than 35% held offshore...

  7. It would be interesting to know whether any such consultations took place.

  8. Also be interesting to know what assets private insurers hold, how much is mandated and how much it varies.
    For example I imagine US based insurers would still hold plenty of US treasuries. (Of course it's a much deeper market)

  9. EQC has interesting materials here:

    EQC can't redistribute its fund across asset classes beyond certain ranges without consulting Treasury and Minister of Finance.

    Current makeup is:
    - 27–33% global equities
    - - of which 30–50% passive
    - 55–73% NZ Fixed Interest (NZG and RBNZ only)
    - 0–12% NZ cash ($250m max).

    Note no NZ equities at all. Before 2001, it was all fixed interest and cash, all the time.

    EQC's liability profile means its fund has to be (a) highly liquid at all times, and (b) highly conservative.

    If I'm reading the documents right, they're expected to go for just a 1/30 chance of a 2% loss in any year (!), and they're shooting for a minimum equity risk premium of just 1% (!). Their fund managers must be drawn from the pool of the most boring people on Earth.

    My guesstimate answers to your questions:
    (1) very tight liquidity and risk parameters make hedging for forex risk sufficiently costly that weighting portfolio heavily to overseas not worth it.
    (2) a fair proportion of NZ fixed interest assets need not actually be offloaded in initial stages (NZ cash assets act as a buffer), because payouts from the fund take time. Perhaps the fund could ride out the initial hit to the value of these assets (or even for a bunch of them, depending on terms, wait for maturity).

    Maybe not totally persuasive, but seems plausible to me. Then again, I'm totally not a finance guy -- my grasp on this stuff is tenuous at best. Take it all with a grain of salt.