LA Fires Do Not Pose A Systematic Risk Via Insurance Companies and Banks--For Now At Least

Weston Nakamura, a trader and independent analyst whose newsletter I recommend for its timely and excellent BOJ and Yen commentary, gave an alternative explanation for last week’s rising yields. He claims that insurance companies, particularly the big three Japanese non-life insurance companies, were forced (and continue) to fire-sale their assets, primarily bonds and equities, to pay out claims due to the LA fires. He also refutes the claim that Trump tariffs and the UK budget are suitable explanations.

Weston notes that his idea is pure speculation and could be wrong. This is why I took the chance to share my view—not that I think I am 100% right, either. Weston introduced me to the idea that as a trader you are in the job of mitigating risks and you have to consider even the off-chance scenarios and plan accordingly (see the post linked at the end). And you have to be bold enough to share (and even trade) your view even if you are the only one who believes in it.

Let me synthesize the competing arguments and give my view.

He first argues that where JGBs lead, other yields follow. JGBs fell while US and UK yields rose in the days leading to the election. On election night, JGBs rose sharply and remained elevated while US and UK yields stalled and then drove lower before taking a directional hint from JGBs. And on the day the LA fires broke out, JGBs rose sharply before their US and UK counterparts followed suit.

He observes:

He correctly notes that JGBs careened lower before US election night while the rest soared. But this dynamic happened right after the UK budget. If you recall, at the time of the announcement, the UK bond market panicked and as you can see, gilts rose sharply to trade at a premium to US yields, which is not supposed to happen because UK pensions duration-match their liabilities creating a consistent demand for gilts. This panic caused a flight to safety into JGBs. After it became apparent that Trump would win the election, all three bond yields rose sharply to anticipate his inflationary policies, but at the same time, the market digested Reeve’s budget and gave her the benefit of the doubt. There was a lot of uncertainty about the implementation of Trump’s policies, including about the timeframe, and both the Fed and BOE cut rates a few days after by 25bps. The net effect of this two factors is that the flight-to-safety into JGBs was reversed, and US & UK yields moved lower. Since the BOJ had held rates on Oct 30, JGBs moved sideways and as other rates declined. Then, going into the December meetings, the market thought that the first Fed cut was a mistake and that they should not cut again given the mounting inflationary pressures. On Dec 18th, they cut by 25bps again resulting in an immediate spike in US and UK yields that was the beginning of another leg higher. This time, JGBs did not follow until much later. So no, JGBs do not lead US and UK yields. As to why JGBs later followed the US and UK yields, a lot has to do with the global term premium. Further evidence of this is that gold also soared since Dec 18th.

As I mentioned in the weekly market report, the UK’s bond market woes are a symptom of the rise in the global term premium. Except a poor 30yr auction caused panic. If you look at the first chart, you can clearly see that on Wednesday 8th, the day of the auction, gilts rose above US yields, signaling increased concern about UK fiscal situation. This is because higher yields are expected to be more problematic for the UK than the US i.e. UK’s term premium is higher. Even now, after a lower inflation number two days ago and a good GDP figure today, gilts are at a premium because bond vigilantes still want a response, while Reeves is being coy and un-assuring.

So that puts both arguments—Trump’s potentially inflationary policies and the UK’s fiscal dilemma—back on the table. Back to Japanese insurance companies.

Weston shows that stock prices for Tokio Marine, MS&AD and Sompo co-move with JGBs, which move before UK and US yields. He shows that when the news about the fire broke out, it was morning in Japan, and late afternoon in California. He blames part of the delay in UK gilts cash trading opens when Tokyo is closing, but argues that US treasuries trade round the clock so they would be open in Tokyo hours.

But, UK bond auctions take place at 10:00 am London time, and if you look at the hourly candle charts, you can see the spike was right after the poor 30yr auction results, not at open (8:00 am).

This shows that yes, investors sold gilts due to fiscal concerns, not the LA fire. The spike in US yields later in the week is a reaction to NFPs.

This does not mean the theory about insurance companies is wrong—just that Weston’s account of the dynamics doesn’t fit. The problem is that first, he shows that JGBs lead other yields, but this isn’t true as I’ve explained, and then he shows that the 3 Japanese insurers co-move with the 10yr JGB. This part could be right. The Japanese firms could have immediately started selling their cross-shareholdings (shares in each other’s companies), and selling their JGBs. If you include algos and investors dumping the stocks since they were aware that the companies were exposed, then the down moves in the stocks make sense.

He also notes that the equities sell-off on Tuesday could be related. My problem with this idea is that the scale doesn’t fit. When the stock market fell on Tuesday, hundreds of billions in market cap was lost, which dwarfs the amount of potential claims due to the fires (currently estimated at $30bn). The three Japanese insurers are only exposed to 3% of that.

So even if they are already making payments, they are not large enough to account for the market moves, and we will only know if they pose a systematic threat later.

To be sure, insurance companies are sweating bullets and have to write some big checks.

And to make matters worse, California insurance companies had no way of estimating the potential damage because a law required them to use past information rather than forecasts.

The problem this poses is that the losses are bigger than they expected and some insurers may go bankrupt. Even home owners who didn’t have insurance have an insurer of last resort, knows as FAIR who will have to write big checks.

The jury is out on whether or not this will pose a systematic risk.

Another concern is a rise in non-performing loans from the mortgages and businesses within the affected areas. It is difficult to assess the extent this will be a problem since non-performing loans take 90 days to reflect in balance sheets.

Side note: Apropos systematic risk, quantitative finance does not yet have competent systematic risk measures or models. We can’t even agree on the definition of systematic risk. So even if a systematic risk emerges, we will find out too late.

As always, feedback and criticism are welcome.