2018 shareholders letter

[wachtwoord]

I agree the tone of the letter is infantile. I'm sure Watsa isn't infantile but coming across as such is not good.

I also believe Watsa's macro view is highly flawed. This wouldn't be an issue if he would act macro agnostic however he made huge bets in the past and even though he claims to have learned from the past I see him comment on Bitcoin which is a macro thing and quite clearly outside of his circle of competance. He seems to overly rely on his own ability on matters he knows little about which is a large risk going forward.

Finally I agree that the 3rd world insurance businesses look like crap. What's the plan there?

[StubbleJumper]

5) Why is there still conflicting messages about hedging?  On the CC, the CEO mishandled a discussion about a potential drawdown in equity markets.  Prem's letter went to great lengths to communicate that there would be no more equity hedging because FFH learned its lesson.  It also observes that US economic growth has been strong, rates have bumped up, inflation looks like it's bottomed out, there's a long runway, etc, but despite those clear and obvious conditions, FFH will continue to hold CPI derivatives worth a notional $114 billion.  About this, I am perplexed.  It's only a $25m market value, but if you are trying to communicate that you are moving on from hedging mistakes in the past, why not liquidate this position.  That large notional value demonstrates that FFH was only speculating on inflation to begin with (ie, if you were hedging you would choose a hedge ratio and then buy your protection...the CPI derivatives exceed FFH's combined assets and gross revenue for a year).  If Prem and Brian have changed their perspective about world markets (which is a legitimate thing to do, and there's no shame in changing your mind), just sell the damned things and recoup the $25m, which happens to be about $1/share of capital.

Haven't read the letter yet, but thanks for your summary - useful.

I'm not sure relating the notional to revenues or assets is useful. They'd only make the notional if absolute CPI went to zero, IIRC, and that seems unlikely! They could have made a couple of billion, maybe more in a depression, but nothing like the size of the notional. My major complaint is not that they hold this but that they should have structured their hedges this way: deep out of the money derivatives that offer outsized gains on low probability outcomes for (relatively) low absolute cost.

Okay, I guess the starting point of the conversation should be, "What is FFH trying to hedge against?"  What would be the bad outcome of deflation that requires protection?  I can offer a few possibilities: 1) FFH's fixed interest debt increases in real terms as a result of deflation, 2) FFH's equity portfolio might be adversely affected by deflation, 3) the collectability of amounts from reinsurers might become dubious, 4) corporate bonds could become shakey, 5) other?

On the liability side, deflation might actually result in reserve releases because presumably IBNR would decline?  Deflation would be good for the real value of their sovereign bonds and possibly their munis.  What else?

So net it out: what's their net exposure?  What kind of hedge ratio should be selected for that notional exposure (somewhere between 0% and 100%)?

I thought I was being rather charitable when I compared the $100+ billion to their total assets.  If you assume that the entire asset base would be adversely affected by inflation with no offsetting benefit on liabilities, and if you were so risk averse that you wanted a 100% hedge ratio, you need what, ~$64 billion notional?  Being even more charitable, assume that a year of revenue would all be adversely affected with no offsetting benefit from expenses, that would be another $18-ish billion?  So in my wildest dreams, that wold be ~$80 billion notional protection required?

Hedging is hedging.  Speculation is speculation.  So which is FFH currently doing, and does it mesh in any way with Prem's broader macro observations in the letter?

Re the other insurance operations, no, you're not alone, although several of these are in places where you can earn high real interest rates in fixed income with relatively little risk (e.g. Brazilian sovereign bonds). That transforms the economics of a 98% CR. Also, some of these businesses will probably benefit from operating leverage as they scale.

Sure, that's the theory, but what do you make of page 59 in the AR (or for that matter, the Asia breakdown on page 115)?  Did we get a fair return on the capital that FFH has deployed? Adjusting for the risk of holding assets in shit-hole countries which do not always have a strong legal system, low levels of corruption, or stable central bank policy, are you happy with what you see on page 59?  What kind of return would be fair for the risks involved with shit-hole assets?


SJ

[Spekulatius]

I have reduced my shares before the annual report came out, and sold most of my remaining shares, except a tracking position after reaiding this.

Reasons
1) share solution ( ~2M more shares). This isn’t Teledyne.
2) continued book value losses.
3) too many crappy investments. FFH is really too complex for its size and it’s not working.

I recycled some funds into BRK last Friday, but have further thinking to do how to reinvest the proceeds.

[rb]

Re the other insurance operations, no, you're not alone, although several of these are in places where you can earn high real interest rates in fixed income with relatively little risk (e.g. Brazilian sovereign bonds).

If that's the case and you feel so confident then why don't you just go and buy a shitload of Brazilian bonds instead?

[TwoCitiesCapital]

I have reduced my shares before the annual report came out, and sold most of my remaining shares, except a tracking position after reaiding this.

Reasons
1) share solution ( ~2M more shares). This isn’t Teledyne.
2) continued book value losses.
3) too many crappy investments. FFH is really too complex for its size and it’s not working.

I recycled some funds into BRK last Friday, but have further thinking to do how to reinvest the proceeds.

I've also reduced my shares, but it was basically because my thesis didn't play out. I loaded up on Fairfax expecting two things to happen:

1) Interest/Dividend income would dramatically increase over the following 2-3 year period due to rising rates which would lead to share price appreciation
2) Significant buybacks at prices from $450-550 USD would likely be accretive if income was to rise dramatically

I sold most of the shares that I had purchased based on disappointing outcomes in both regards.

10-year rates rose to 3.2% prior to falling back down to 2.6%. With low inflation and limited pressure in the near/mid-term from rising front-end rates, it's hard for me to make a case for rates to rise as significantly as originally anticipated. I'm not saying Fairfax should've played the short-term game, but they did miss an opportunity to lock in longer-term rates and the opportunity set for increasing interest income in the near-term appears challenged.

Further, buybacks have been less than I anticipated based on Prem's first comparison to Teledyne...and have been further diluted by compensation and share issuance that were unanticipated at the time of purchase.

I still some shares and would be willing to add in the low 400s near book value, but there doesn't appear to the be the catalyst for anything to change with the company or help them achieve the 15% ROE that I initially anticipated.

[petec]

Re the other insurance operations, no, you're not alone, although several of these are in places where you can earn high real interest rates in fixed income with relatively little risk (e.g. Brazilian sovereign bonds).

If that's the case and you feel so confident then why don't you just go and buy a shitload of Brazilian bonds instead?

The childish answer is: because I already have a shitload of Brazilian equities.

The better answer is: because as you well know, a float-levered insureco with a sub-100% underwriting margin is going to return more than a sovereign bond even if 100% of its float is invested in that same sovereign bond. Fairfax's Brazil operation has underwriting profits, and has growth potential since insurance is underpenetrated in Brazil and the country is entering a cyclical upswing and possibly also a disinflationary boom (if the government gets its reforms done).

The majority of the underwriting losses in Fairfax's Latin American operations came from Argentina, which (unlike Brazil) is a hyperinflationary basket case - but it is undergoing some very promising reforms. Time will tell.

[petec]

I agree the tone of the letter is infantile. I'm sure Watsa isn't infantile but coming across as such is not good.

I also believe Watsa's macro view is highly flawed. This wouldn't be an issue if he would act macro agnostic however he made huge bets in the past and even though he claims to have learned from the past I see him comment on Bitcoin which is a macro thing and quite clearly outside of his circle of competance. He seems to overly rely on his own ability on matters he knows little about which is a large risk going forward.

Finally I agree that the 3rd world insurance businesses look like crap. What's the plan there?

Out of interest do you think Watsa's style has changed or did you not like it from the start (whenever that was, for you)?

Watsa has as much right to comment on bitcoin as anyone else. Bitcoin can only be understood through two lenses: 1) crowd psychology and tulip bubbles and 2) fiat currency collapse. The first - which is the one he used - is well within his circle of competence as a value investor.

My framework for the third world insurance businesses is this: if one or two of them find that sweet nexus you occasionally get in insurance where a superb manager meets a superb opportunity, then you've seeded the next ICICI Lombard or First Capital. My guess is that's the game plan, but it's only a guess. The opportunity stems from the fact that in most of these places insurance will grow faster than GDP, because it is currently underpenetrated.

[petec]

Okay, I guess the starting point of the conversation should be, "What is FFH trying to hedge against?"  What would be the bad outcome of deflation that requires protection?  I can offer a few possibilities: 1) FFH's fixed interest debt increases in real terms as a result of deflation, 2) FFH's equity portfolio might be adversely affected by deflation, 3) the collectability of amounts from reinsurers might become dubious, 4) corporate bonds could become shakey, 5) other?

On the liability side, deflation might actually result in reserve releases because presumably IBNR would decline?  Deflation would be good for the real value of their sovereign bonds and possibly their munis.  What else?

So net it out: what's their net exposure?  What kind of hedge ratio should be selected for that notional exposure (somewhere between 0% and 100%)?

I thought I was being rather charitable when I compared the $100+ billion to their total assets.  If you assume that the entire asset base would be adversely affected by inflation with no offsetting benefit on liabilities, and if you were so risk averse that you wanted a 100% hedge ratio, you need what, ~$64 billion notional?  Being even more charitable, assume that a year of revenue would all be adversely affected with no offsetting benefit from expenses, that would be another $18-ish billion?  So in my wildest dreams, that wold be ~$80 billion notional protection required?

Hedging is hedging.  Speculation is speculation.  So which is FFH currently doing, and does it mesh in any way with Prem's broader macro observations in the letter?

Re the other insurance operations, no, you're not alone, although several of these are in places where you can earn high real interest rates in fixed income with relatively little risk (e.g. Brazilian sovereign bonds). That transforms the economics of a 98% CR. Also, some of these businesses will probably benefit from operating leverage as they scale.

Sure, that's the theory, but what do you make of page 59 in the AR (or for that matter, the Asia breakdown on page 115)?  Did we get a fair return on the capital that FFH has deployed? Adjusting for the risk of holding assets in shit-hole countries which do not always have a strong legal system, low levels of corruption, or stable central bank policy, are you happy with what you see on page 59?  What kind of return would be fair for the risks involved with shit-hole assets?

SJ

I agree with your reasoning, but I think it is based on faulty facts:

1) Back when they were worried about debt bubbles collapsing Fairfax looked hard at how insurers performed in the great depression. They discussed this in investor meetings (can't remember if they discussed it in letters/calls). Obviously equities collapsed (= FFH book value virtually gone). More surprisingly underwriting profits disappeared too as cash-poor insurers desperately competed to write premiums in the face of collapsing demand. The result was not that there were offsetting drivers in insureco P&Ls, but a total meltdown which most companies did not survive. I think that fear is what drove their entire hedging programme - both equity and deflation swaps.

2) As I understand it if you buy $100bn notional and CPI drops 10% below strike, you make $10bn. To make the full notional amount of $100bn, CPI would have to fall 100%, which is unlikely since it implies that the CPI basket of goods would cost nothing and the value of money would be infinite. Given that Fairfax regularly highlighted that CPI dropped 17% in both the GD and in Japan post-bubble, I assume that's the kind of thing they were worried about if things got really bad. If so, $110bn notional protected at most $20bn of assets. In fact I think it was quite a bit less because IIRC Fairfax bought the swaps below strike, meaning you needed several % deflation before you made money. In more likely scenarios, upside was never going to be more than $2-3bn. So under no realistic assumption did they exceed a 100% hedge ratio.

I don't really care what they do with a $25m position, so I won't argue on that one.

Haven't got to the AR yet so can't answer your last question.

[Spekulatius]

I have reduced my shares before the annual report came out, and sold most of my remaining shares, except a tracking position after reaiding this.

Reasons
1) share solution ( ~2M more shares). This isn’t Teledyne.
2) continued book value losses.
3) too many crappy investments. FFH is really too complex for its size and it’s not working.

I recycled some funds into BRK last Friday, but have further thinking to do how to reinvest the proceeds.

I've also reduced my shares, but it was basically because my thesis didn't play out. I loaded up on Fairfax expecting two things to happen:

1) Interest/Dividend income would dramatically increase over the following 2-3 year period due to rising rates which would lead to share price appreciation
2) Significant buybacks at prices from $450-550 USD would likely be accretive if income was to rise dramatically

I sold most of the shares that I had purchased based on disappointing outcomes in both regards.

10-year rates rose to 3.2% prior to falling back down to 2.6%. With low inflation and limited pressure in the near/mid-term from rising front-end rates, it's hard for me to make a case for rates to rise as significantly as originally anticipated. I'm not saying Fairfax should've played the short-term game, but they did miss an opportunity to lock in longer-term rates and the opportunity set for increasing interest income in the near-term appears challenged.

Further, buybacks have been less than I anticipated based on Prem's first comparison to Teledyne...and have been further diluted by compensation and share issuance that were unanticipated at the time of purchase.

I still some shares and would be willing to add in the low 400s near book value, but there doesn't appear to the be the catalyst for anything to change with the company or help them achieve the 15% ROE that I initially anticipated.

I agree on all counts. I think it’s fair to say that the thesis as most of us envisioned it a year ago didn’t play out and is unlikely to play out in the near term future.

[vinod1]

StubbleJumper makes very good points. However, I think shorting and CPI bets are better understood from the perspective of Mr. Watsa the individual.

Rewind back to the 2007-2009 period. Fairfax reported massive gains on the CDS portfolio, and many of us who have loaded up on Fairfax and its subs felt great at that time and patting ourselves on our backs for this. 

How would this have felt for Prem? After all, we Fairfax investors were on a high just from taking advantage of Fairfax gains. He is one of the handful of investors who came out with billions of dollars of gains directly during the period. For 3 years book value exploded and it is hardly a stretch to think he must have felt like a genius, especially when every other company is wallowing in misery.

I can recall a period like that, when my LEAPS on BAC paid off big time, then after a couple of months of research on O&G companies, invested in Sandridge Energy LEAPS and they paid off like 5x in a few short months. I was on a high and contrary to my normal behavior invested in HP LEAPS with very little research. Fortunately, it is only 0.1% of portfolio and to make a long story short, the HP LEAPS went to zero the moment Leo Apotheker did a deal with Autonomy.

I think the "rush" from the 2007-2009 gains might have been a contributing factor. It was never about hedging.

If you are really concerned about Great Depression type scenario, do you invest in crappy companies like Blackberry, Sandridge Energy, Greek/Irish banks, restaurants?

Further look and what he has done on the business side, he went ahead and bought a bunch of other companies.

Is this what you expect when you are preparing another Great Depression?

(Hat tip to UCCMAL for pointing these out in one of the earlier discussions.)

One more thing. Even when you have hedged 100% of the equity portfolio, you promise to investors 15% annual returns? There is a snowball chance in hell that you would reach that objective without a major market crash.

The only logical explanation that fits the facts is that Prem is expected a market crash and the S&P 500 shorts and CPI portfolio is a market call. Pure and simple. It is a market call that did not work out.

All the explanations - avoiding 1-100 year event, hedging, etc. do not fit the facts.

Prem is a wonderful person from everything that I gather. He has build a multi billion dollar company from scratch. Nothing can take these things away from him.

But as investors we have to separate Prem the wonderful person from Prem the CEO. He got the market call wrong. It happens. The problem is not being able to acknowledge it. Calling it hedging is just plain wrong.   

Vinod