Comments and Observations about A/R

[StubbleJumper]

I broadly agree. The share buyback doesn’t make me scratch my head because I always read that as a long term thing. Riverstone UK does a little - not sure why it’s better placed to grow with OMERS as a partner, which is what Prem says - but that negative is offset by my surprise at how much it’s worth and we have to consider the possibility that the motivation for the deal may simply have been to surface value and get the BV mark.

But what does really piss me off is that Prem keeps banging on about how well capitalized the subs are while not mentioning that they had to squirt equity into some of them during the year. Just feels like a lie.

My take is that they sold Riverstone because they needed the cash, and even when added to the divvies from the subs, the $560m that they are getting probably won't be enough to fund the holdco's activities for 2020.  So, until they float a debt offering, they seem to be reliant on that revolver -- IMO, it is not a great idea to have your operations reliant on operating credit because usually it is accompanied by 25 pages of covenants which means that your operating credit can get pulled if something weird happens (ie, debt:equity soars due to stock market crash or asset write-downs, EBITDA drys up for some reason, etc).  A large revolver is a nice tool to keep around for emergencies or to opportunistically exploit opportunities (ie, suddenly buy a $1B sub or buy $1B of convertible notes on short notice).  But, once you draw on it, the preferred approach would be to replace it by floating a bond issue to obtain long term funds for which the indentures are less restrictive than the 25 pages of revolver covenants.

My point about the buybacks was about both tone and strategic direction.  Pull up the annual letter from March 2018 and read it again.  There's the usual rah-rah, we're awesome, and more than a few humble-brags.  But the message was that the masterpiece was pretty much painted, the acquisitions were about done and buybacks would be the way forward -- heavens, Prem even had the temerity to trot out Henry Singleton's name, as if to suggest that FFH would initiate a long-term process to buy back ~90% of its shares!  In that letter, he shared the usual "Financial Position" Table, but for the first time that I can recall he gave us the song and dance about dividing the insurance companies from the non-insurance companies and differentiating the debt by claiming that the non-insurance debt is non-recourse, so somehow that made it okay (more on that later).  But, take a look at that same table from Friday night.  Using Prem's mental accounts approach, compare the Net debt/Equity Ratio from two years ago to that from Friday.  That ratio has risen from 16% to 25.5%.  Seriously, pull up the those two letters and read them in succession and tell me that you are not left scratching your head. 

So the announced strategy in 2018 was to temper growth through external acquisitions, solidify the balance sheet, and return capital to shareholders through buybacks, but what we instead saw was injections of capital into Toys R Us, Carillon, AGT, Stelco, Fairfax Africa, Seaspan, etc, in addition to the expected investment in Brit (at least the Brit injection was consistent with the announced direction).  From the outside it is difficult to assess where they sourced the cash for those moves -- it could have been subsidiary cash or it could have been holdco cash, but it doesn't really matter.  The fact of the matter is that Prem's behaviour as a serial acquirer does not seem to have changed one whit, and FFH has continued on a debt-financed growth spree for the past two years.

This is, unfortunately, a little reminiscent of the past.  Debt-financed growth is great when it works and disastrous when it doesn't.  Prem's protestations about non-insurance debt being non-recourse are not re-assuring.  At least in '03 or '04 he postured-up and publicly declared that ORH debt was FFH debt and it would be managed and treated as such, irrespective of whether ORH and the other FFH subs were cross-collateralized.  Now we have this strange weasle-ish notion that FFH would walk away from a troubled non-insurance sub, so we don't need to bother counting that debt.  But, on that, I call bullshit.  If FFH ever walks away from any debt, insurance or non-insurance, who the hell would ever trust them to honour an insurance policy?  It's *all* FFH debt, irrespective of whether it is cross-collateralized, full-stop.

So, yes, I am just a little bit uncomfortable that quality of earnings has been dubious for the past two or three years, debt has grown, some of the subs seem to be capital constrained, a portion of Riverstone is being sold despite Prem's assertion from two years ago and this year that it is an excellent vehicle for organic growth, and FFH seems to be using operating credit at the holdco.  Does this seem like well planned implementation of strategy to you?  It strikes me as a bit of a chaotic and ad hoc approach to managing a $70B enterprise.  We've been here before.


SJ

[ValueMaven]

Trading at x0.78 of BV?  What am I missing? 

[Cigarbutt]

...
5) The Loss Triangles - it was nice to get updated loss triangles because they are essential to better understand what's going on with UW (see page 79 of the AR).  As I have observed on several occasions over the past year, favourable development has been chronically ridiculous, regularly clocking in at 10% of reserves.  All of that came to a screeching halt in 2019, with only $100m of redundancies on $29B of reserves.  The decline has actually been a bit more gradual than it appears, if you strip out the impact of currency translation and getting slapped in the head once again by APH (it's really a "gift" that keeps on giving...how much of the APH liability is related to the TIG and C&F acquisitions?).  Stripping out those two factors, the change in favourable development isn't so bad.  Once again, this year I would note my disappointment about not getting the loss triangles for each of the major subs....somehow I suspect that C&F's is a bit of a shit-show, but we'll never know.
...
SJ

Thank you for the salient points. Below are some thoughts on reserves.
Reviewing the annual report is a reminder of how FFH has built amazing zones of strength while keeping areas of vulnerability. I guess that's why the stock is hard to handicap, especially at this juncture.
Insurance and reinsurance remains the backbone of the business.

Favourable loss reserve development (2010-2018, without the currency effect)
336.9  432.1  1,136.9  1,383.7  1,429.0  1,549.3  907.0  608.2  166.0
FFH does not disclose that but it’s possible to reconstruct year by year development for each calendar year business. A potential problem is the embedded currency effect which was quite positive in the 2018 annual report (2017 reserves column listed below).
The following is unaudited:
Reserve development (numbers in (    ) are unfavorable, as reported end of year+1)
                2016     2017     2018
2013         74.3      45.0      3.5
2014        175.9     91.4     25.9
2015         40.5     296.0   144.1
2016       (69.8 )    67.2     18.6
2017                    483.2   (37.3)
2018                                 84.7

-The trend is variable but down.
-The 2017 reversal is significant.
It’s conceivable that FFH reports net negative adverse development as early as next year..

@ValueMaven
With marked-to-market rules and perhaps a transitional new normal for volatility, book value may be a moving target.

[StubbleJumper]

...
5) The Loss Triangles - it was nice to get updated loss triangles because they are essential to better understand what's going on with UW (see page 79 of the AR).  As I have observed on several occasions over the past year, favourable development has been chronically ridiculous, regularly clocking in at 10% of reserves.  All of that came to a screeching halt in 2019, with only $100m of redundancies on $29B of reserves.  The decline has actually been a bit more gradual than it appears, if you strip out the impact of currency translation and getting slapped in the head once again by APH (it's really a "gift" that keeps on giving...how much of the APH liability is related to the TIG and C&F acquisitions?).  Stripping out those two factors, the change in favourable development isn't so bad.  Once again, this year I would note my disappointment about not getting the loss triangles for each of the major subs....somehow I suspect that C&F's is a bit of a shit-show, but we'll never know.
...
SJ

Thank you for the salient points. Below are some thoughts on reserves.
Reviewing the annual report is a reminder of how FFH has built amazing zones of strength while keeping areas of vulnerability. I guess that's why the stock is hard to handicap, especially at this juncture.
Insurance and reinsurance remains the backbone of the business.

Favourable loss reserve development (2010-2018, without the currency effect)
336.9  432.1  1,136.9  1,383.7  1,429.0  1,549.3  907.0  608.2  166.0
FFH does not disclose that but it’s possible to reconstruct year by year development for each calendar year business. A potential problem is the embedded currency effect which was quite positive in the 2018 annual report (2017 reserves column listed below).
The following is unaudited:
Reserve development (numbers in (    ) are unfavorable, as reported end of year+1)
                2016     2017     2018
2013         74.3      45.0      3.5
2014        175.9     91.4     25.9
2015         40.5     296.0   144.1
2016       (69.8 )    67.2     18.6
2017                    483.2   (37.3)
2018                                 84.7

-The trend is variable but down.
-The 2017 reversal is significant.
It’s conceivable that FFH reports net negative adverse development as early as next year..

@ValueMaven
With marked-to-market rules and perhaps a transitional new normal for volatility, book value may be a moving target.

Thanks for the accident year transposition.  I hope your concern about 2017 and onward is wrong.  FFH has gone from providing copious detail about reserve development to providing very little.  I guess that's one of the challenges of reporting on such a large enterprise.  If you don't want a 1,000 page AR, you need to chop some material.


SJ

[petec]

Trading at x0.78 of BV?  What am I missing?

The fact that BV has dropped a lot - check out the Eurobank and Atlas Holdings share prices.

Now, those stocks are cheap so plenty of value has opened up. But it’s arguably in the holdings, not Fairfax.

[TwoCitiesCapital]

Trading at x0.78 of BV?  What am I missing?

The fact that BV has dropped a lot - check out the Eurobank and Atlas Holdings share prices.

Now, those stocks are cheap so plenty of value has opened up. But it’s arguably in the holdings, not Fairfax.

And that interest rates are below 1% all the way down the curve meaning getting a 10-15% ROE is a pipe-dream unless if the insurance markets just go gangbusters.

[petec]

Trading at x0.78 of BV?  What am I missing?

The fact that BV has dropped a lot - check out the Eurobank and Atlas Holdings share prices.

Now, those stocks are cheap so plenty of value has opened up. But it’s arguably in the holdings, not Fairfax.

And that interest rates are below 1% all the way down the curve meaning getting a 10-15% ROE is a pipe-dream unless if the insurance markets just go gangbusters.

Agreed. Although *arguably* one ought to be looking at the ROE/risk free rate spread when valuing a stock.

[jfan]

Just a simple question:

1) When determining the statutory surplus, is it almost equivalent to Common shareholder's equity + preferred shares + minority interest?

Just a simple observation:

Referring to their annual report's total return on investment portfolio footnote, it appears that they under-perform their 7% investment return hurdle rate 39% of the time over the past 33 years.

If that is true, that means under-performing 5 years in a row, would put it at less than a 1% probability.

[TwoCitiesCapital]

Trading at x0.78 of BV?  What am I missing?

The fact that BV has dropped a lot - check out the Eurobank and Atlas Holdings share prices.

Now, those stocks are cheap so plenty of value has opened up. But it’s arguably in the holdings, not Fairfax.

And that interest rates are below 1% all the way down the curve meaning getting a 10-15% ROE is a pipe-dream unless if the insurance markets just go gangbusters.

Agreed. Although *arguably* one ought to be looking at the ROE/risk free rate spread when valuing a stock.

I think that makes sense in theory, but in practice the swings in the "risk free rate" are far too volatile to be basing my values off of.

I don't think the actual values of companies have swing as dramatically upward as the 10-year going from 3.25% down to 0.3% might imply. Nor do I think those same companies lose dramatic value if it goes back to 3.25% in the next 18-24 months. Even using Fed funds rate instead, pretty dramatic swings.

So either an absolute rate of return or some longer-term moving average need to be used as your risk free rate.

Also, I'd say the swings with Fairfax should be more muted relative to the risk free rate as well given its future earnings are somewhat counter-cyclical (i.e. interest earnings dramatically increase as your discount rate is rising offsetting some of the impact. Vice versa as rates are falling).

[Cigarbutt]

Just a simple question:

1) When determining the statutory surplus, is it almost equivalent to Common shareholder's equity + preferred shares + minority interest?

2) Just a simple observation:

Referring to their annual report's total return on investment portfolio footnote, it appears that they under-perform their 7% investment return hurdle rate 39% of the time over the past 33 years.

If that is true, that means under-performing 5 years in a row, would put it at less than a 1% probability.

For 1), statutory surplus is determined by state regulators who typically use a risk-based capital framework (similar to banks) to reduce the value of certain elements (and increase the margin of safety for the policyholder) of the balance sheet, as reported. The discounts vary and depend on the perceived level of risk. FWIW, I've been looking at a few insurers who carry a heavy load of BBB rated corporate bonds (not the case for FFH). An interesting feature is that, in the event of a recession, on top of the decrease in market value for the bonds, surplus capital gets a double whammy because the discount factor is higher for downgraded securities.
For 2), your statistical appreciation of forward returns is interesting and is in line with the idea of reversion to the mean, which has been a significant long-term feature at Fairfax (investment strategy, seven lean years analogy etc) but I wonder if such an approach is satisfactory on a forward basis as the investing environment has changed and the Fairfax investment recipe has been changing (some aspects dramatically so) so the future may not be correlated to the past. I think I read you're an MD and the following statistical "joke" came to mind when reading your post. There's this surgeon who comes to the patient waiting to be rolled in and explains that the death risk with the procedure is 1 in 2 but that the patient should not worry because the previous patient did not make it.