A lengthy screed on Tower Limited
I’m revealing my undisclosed 10th purchase a little sooner than expected: It’s Tower Insurance Limited (ASX: TWR). I expect a 3-5 year holding period from here, subject to business performance. Conditions are also not quite as favourable as I had hoped for, so I have taken a smaller position with a view to increasing if a better opportunity arises.
This post is a longer follow-up to my Buy thesis on Tower. There are some overlaps, because I want to have all of my sources in one place.
One thing that you should keep in mind is that I wrote this post at the time of the purchase, so some comments – about management’s shareholdings, or Goldman Sachs’ fees, for example – are outdated. I have kept them there because it reflects the information I had at the time I purchased in late 2017. Do your own research.
Tower Insurance
- A$212m market capitalisation (337.3m shares on issue @A$0.63 share price)
- Underlying profit after tax of NZ$18m (A$16.5m; P/E of ~13x)
- NTA of NZ$0.90/share as of 30 September 2017
Tower is a New Zealand insurer that spent the last 10 years selling off all its businesses and returning capital to shareholders only to get caught short of capital following the Christchurch earthquake in 2011. Now it’s stuck with antiquated IT systems and had to raise substantial additional capital from shareholders. Reading between the lines, you can see historical hints of underperformance, technological problems, and underinvestment in the business at the very same time as management is focusing on their dividend and returning capital to shareholders.
That was before the appointment of ex-Territory Insurance Office (TIO) CEO Richard Harding in 2015, who I think has done a solid job turning things around (ongoing collapse in the company’s share price notwithstanding). Prior to Tower, Harding was the CEO of TIO, the NT government insurer, for around 6 years. He led it through a period of improvement in operating profit prior to its sale to Allianz in 2014.
Old TIO annual reports can be found here: http://www.territorystories.nt.gov.au/jspui/handle/10070/242199
Like other insurers, Tower still has unresolved claims from the Christchurch earthquake in 2011. IAG discreetly raised capital for this in 2016 but Tower postponed doing so due to two takeover bids. Bids for Tower were launched at $1.17/share (by Fairfax of Canada) and $1.30/share (Vero/Suncorp Group) in 2016. The first bid was invalidated by the higher bid, and the Vero bid was rejected by the NZ Commerce Commission because it would result in Suncorp and IAG having a duopoly.
10foot goes activist
As part of my initial look at the stock, I spoke to some of the shareholders in Tower. They weren’t keen on saying much (as I am anonymous) but I got the impression that they had basically capitulated and wanted to sell the company in the takeover and wash their hands of it. I also sent an early version of this post to some small cap managers and their general response was “yeah it’s an insurer, no thanks.” That could mean that this purchase is not smart – be aware that several smart investors turned this idea down – but I took it to mean that there was close to zero interest in Tower and a possible opportunity where nobody was looking, so to speak. I saw that some NZ brokers named it as their top pick for 2018, so it’s probably going straight to zero.
(That’s a joke. Actually there is one brokerage firm that has had fantastic – detailed and practical – coverage of Tower, and that’s Forsyth Barr. They are often quoted in the media links I provide below.)
In another world I might have gone all Bill Ackman on Tower – invest billions, watch it plunge 70% over the next few years, then get sued for insider trading – but as a certified small fry I was able to avoid that temptation. I do hope to get an interview with CEO Harding for this blog and I will be approaching Tower to that end very shortly.
The thesis:
Tower has a lot of unrecognised opportunities in its core business and today’s price – ~13x earnings and ~1x book is not the right one.
For example, until recently Tower did not have an online sales option, did not have a preferred supplier network for vehicle repairs, and had over 300 policy types that had not been repriced or been updated for ‘years’. The unresolved Christchurch earthquake claims were also weighing down the business and the share price.
Following reinvestment, Tower should see significant improvements in every area of the business, especially operating costs and risk pricing. I also reckon retention and policy numbers should improve. Tower has been growing policy numbers and reducing management expenses for the last 2-3 years even with its weak business systems, which I think is remarkable given the challenges it has faced. For example, Tower previously had three separate teams for selling, renewing, and updating a policy.
I direct you to this old AGM speech from CEO Harding that sums up everything that’s been happening.
Note this link will download the 2.6mb PDF file directly to your device:
The Tower opportunity:
With low-hanging fruit in every area of the business, expected improvement to earnings should more than compensate for the capital raising over time, and Tower is currently undervalued in my opinion.
Given the extent of the improvements and historical growth in policy numbers, market share, and reduction in costs, and the benefits Tower expects to achieve from these, I estimate fair value for Tower is around A$0.90 following business improvement. I estimate the company may (in time) pay a ~7% dividend at my overall purchase price of $0.55, based on a 50% payout ratio. These estimates are just part of a wide range of possible outcomes so do not rely on them.
The nullification of the Christchurch liabilities (management was going to spin them off, but they didn’t mention this in the recent results) should ease the process of re-rating.
I think the downside on Tower is limited and the upside is meaningful without being outstanding, especially if dividends recommence in the next 2-3 years as I expect. Management said they may recommence dividends as early as FY18. There is reasonable probability of a blue-sky scenario in which Tower claims market share, controls claim costs, and reduces operating expenses further than expected, but I am not counting on that. Further takeover bids, e.g. from Fairfax, are a possibility.
The balance of probabilities
What is the actual chance of Tower succeeding/failing as an investment?
In general terms, I think it is nearly inconceivable that Tower will not be a better company in five years. The improvements are very intuitive, albeit not simple to implement, and have been demonstrably successful at other insurers. However, whether these improvements actually lead to improved profitability and market share is the key question. I obviously think that they will, but Tower is playing catch-up to larger and more sophisticated competitors, which is not generally a winning premise. This is something to keep in mind. There is also the tail risk of catastrophic cyclones etc. I think the risk of a weather-induced bankruptcy is remote following the capital raising, but is definitely an important consideration.
One argument is that it doesn’t really make sense for Tower to be a standalone listed co as it is small and its cost ratio is so high, but I disagree. I think there is definitely a role for a third player in the Kiwi market and if Tower wasn’t there somebody else would take that position. On a non-business front I also think it would be a real shame for an iconic Kiwi brand like Tower (150 years of history) to be sold off to a foreign multinational like IAG or Fairfax.
The transformation process:
Tower currently has 30% of its sales online, up from 9% just 18 months ago. I think something like 70%-80% online sales is a plausible target (who makes a phone call to buy insurance these days?), and Tower says that it is aiming for more than 50%. This should result in lower costs, better service, and wider margins. There is a scoping study underway for the new IT systems and costs are currently uncertain. IT systems are expected to be deployed by the end of 2018, but I think Tower is kidding itself – I’ve never ever seen an IT system delivered on time, on budget, and without problems. Still, the opportunity from a new IT system appears worth the risk:
I never heard anything more about that investor day. (Tower half year prezzo FY16)
If you know anything at all about software, you can clearly see the benefits of pulling it all together into one system instead of currently 4 systems and ‘dozens of ancillary systems’. The AGM speech I linked earlier has comments on the software system too.
I originally budgeted for a capital raising of 100m new shares at $0.50 to $0.60 – what I actually got was 166m new shares at $0.42, worsening the opportunity due to dilution. I blame Goldman Sachs and Tower – but more on that later. Tower has raised enough capital to cover its Christchurch quake liabilities and more besides. It has 99% claims coverage (including Incurred But Not Reported claims) for this event, and there is an 80% to 85% statistical probability that this will be sufficient to cover all claims (because the claims expense is inherently uncertain). The stability of Christchurch claims is a key risk, given multiple years now of ever-escalating claim expenses.
Management also expects to be in a position to upgrade the IT systems, which was a vital part of the thesis. Ultimately I think management took the hard road by raising a lot of capital, and I applaud them for it as it suggests they focused on the long term value of the business. This is despite their lack of personal shareholdings – but more on that later too.
Tower currently has a ~5% insurance market combined share (up from 4.6% in 2014 and 4.7% in 2015) and has been growing GWP about 3%-4% per annum. Market share has been increasing in recent years, despite lacklustre capabilities, although it is obviously well down from historical highs of ~10% around 9(?) years ago. There are fewer ‘hard’ moats/ network effects that would prevent Tower growing its market share, especially since Tower already has some financial adviser networks, banks, and alternate distribution sources such as NZ Airpoints and TradeMe insurance. Tower is over 150 years old, and has top-tier brand recognition on par with that of IAG/Suncorp owned brands.
Insurance is a commoditised product, and two major competitors (IAG + SUN) control around 70% of the NZ market. I think that Tower could probably claim additional market share over time, as it proves itself a viable and attractive proposition for new customers, although the thesis does not depend on market share growth. Anecdotal reports suggest Tower is ahead of other NZ insurers in its focus on digital, which is encouraging to hear, although I have not relied on it. The NZ Commerce Commission recently commented on price coordination behaviour between the two majors and I expect this opens an opportunity for Tower to compete via differentiation of its brand/products/service.
Management/board lack of alignment:
The chairman is one of the more heavily invested board shareholders and he owns less than one year’s salary worth of Tower shares (albeit he may have purchased at a higher price). He also has more than 20 other directorships(!!). CEO Richard Harding is highly paid and doesn’t own a single share. There are no equity incentives in his remuneration agreement. I think this is probably because he was hired for his expertise and may not have wanted to invest with the Christchurch liabilities etc. Even so, I believe this lack of alignment led to the unfortunate terms of the capital raising (see below). I would like to see management start to purchase serious amounts of shares on market. Chairman Stiassny has reportedly done a good job, but equally I believe he should resign some of his other directorships and/or buy significantly more Tower shares. Tower deserves a more heavily invested board and management team.
(The CEO and Chair recently purchased more shares and participated in the cap raising in a modest way)
From the Tower FY17 annual report remuneration report.
Ordinarily these would be grave concerns, but I have formed a high opinion of the CEO given Tower’s progress over the past two years, and I have spoken with long-time Tower shareholders that have corroborated my impressions. Still, these kinds of turnarounds can benefit significantly from execs with the grit (read: threat to their wealth, or reputation if a founder) to go above + beyond.
Also, in my opinion, weak corporate governance at Tower historically led directly to its weakened state, and I think this is something that should be tightened up ASAP now that the business is looking to rejuvenate itself.
As an aside, if I could speak to Tower directly, I would say please stop saying that you are ‘focused on capital management’. The historical focus on capital management led directly to Tower being undercapitalised in the aftermath of the quake. Given the history here and the current situation (e.g. ongoing dispute with Peak Re, see below) I would vastly prefer to see Tower comfortably overcapitalised while the reforms take place. You can always manage capital downwards but you cannot easily replace it, as long term shareholders have discovered to their considerable chagrin.
An error in Tower’s historical financials?
I had a brief worry when I spotted what looked like an error in Tower’s historical claims expenses:
The 2015/2016 numbers are identical to the 2014/2015 numbers. (via FY16/FY17 annual reports)
I subsequently realised it was probably a typo and I contacted Tower IR who confirmed that it was. The data is in the wrong column, i.e., the ‘2013’ column is actually ‘2012’ etc. I believe them and intuitively this makes sense, although the possibility of an ‘error carried through’, for example in a spreadsheet, is one that every would-be investor should consider for themselves. Orocobre famously had 20% of their annual production vanish into the ether due to a spreadsheet error, so it happens.
Anecdotally I have been surprised by how many similar errors I’ve spotted in company reports (not just Tower’s) when I really look for it. I saw a similar thing with Automotive Solutions Group in early 2017 and I have seen over a dozen notable examples since. I guess they are complex documents and vulnerable to errors.
The Vampire Squid strikes again:
Goldman Sachs got an estimated 6% fee off the top of the capital raising. $4.5m of the $70.8m – $6 out of every $100 raised – raised goes to Goldman, and I think this is a black mark against Tower. The market is quite aware of Tower’s antiquated state, but not necessarily of the opportunities for improvement, and I think if presented in the right way, more holders could have been convinced to participate. My belief is that the capital raising was under-marketed and under-priced.
(Subsequently, it turns out that the capital raising was 12% undersubscribed, and with such little liquidity Goldman is going to earn its underwriting fee. I estimate they now own/owned about 4% of Tower)
The price at which capital was raised was also pretty ‘eh’ with NZ$0.42 being a ~40% discount to the last trade price. Vero may have had a role to play here by demanding a greater discount in order to support the cap raise, no doubt trying to average down its very high buy price of $1.40/share.
Still, in my opinion this is the first sign of management’s lack of alignment hurting shareholders – I doubt that a founder would have paid so much to an investment bank or issued these shares so cheaply. A founder may also have more viscerally grasped the need to pound the bricks and drum up interest in Tower. Perhaps Tower did this in New Zealand (where most shareholders live) but I did not see any evidence of marketing on this side of the ditch.
In Tower’s defence, it did a renounceable pro-rata rights issue (huge credit to them for this) and I think in general that the lack of self-promotion is a big plus to Tower. They have done the right thing by shareholders. This company strikes me as an old-school insurer with a focus on risk rather than sales, which is ideal. However, it could really have used a salesman in the lead-up to that cap raise.
While I am not super happy with how this part turned out, I don’t think management is in the wrong – they needed capital and it was a responsible decision to underwrite it. I just would have preferred to see a little more aggression when raising, and a little less going to Goldman. I mean, if fucking Getswift can raise $75m on the thinnest of premises, Tower – established business, already making progress, solid turnaround plan – should have absolutely cleaned up.
It’s Seduction 101: You gotta make them want it.
Dispute with Peak Re
I think this was a pretty shrewd move from Tower. Tower bought expensive reinsurance from Peak Re to cover potential increases in the Christchurch quake claims. It looks as though Tower effectively passed some of their liabilities off to someone else – the reinsurance was subsequently quickly drawn on and Peak Re was not impressed. Tower’s claim is now in dispute. Peak Re hasn’t handed over any money, yet Tower has counted the funds from this claim as part of its balance sheet, leading to risks if there is drawn out litigation or the contract gets void. That would leave a ~$40 million hole.
Tower says it is in the right, and I also struggle to see how Peak Re will wriggle out of it. However (if I understand correctly) if Tower knew with a high degree of probability that its future liabilities were likely to tap that reinsurance, then the contract may be void and Tower would be left with a hole in its balance sheet. Notably Tower hasn’t got the cash out of them yet and as they say, possession is 9/10ths of the law, so protracted litigation here might lead to Tower having to raise more capital or delay paying dividends.
Here is an excellent media article explaining the situation from interest.co.nz.
Here are several more media reports that are well informed and explain the wider industry risks in NZ
I don’t want to go too much into wider industry issues as they require a lengthy discussion. Generally I am of the view that
a) consumers alone can’t affordably insure their homes against earthquakes without something like the EQC. I expect government support to remain via one mechanism or another.
b) risk-based pricing for Tower will give it the ability to better price earthquake/flood prone areas – and also perhaps to price itself out of areas it does not want to insure, if that should become necessary.
c) I am not certain but I believe that Tower has moved away from ‘replacement’ policies towards ‘sum insured’ which should reduce complexity and cost of its claims when future events occur.
The risks:
AUD/NZD exchange rate is a risk but difficult to forecast. It is hard to foresee significant, sustained divergence over the long term. If anything I’d be bearish AUD with China etc.
IT simplification is rarely trouble free but in this case benefits are clear enough to run that risk. Tower’s IT systems are sub-par and long-term shareholders agree that there are opportunities for improvement in every area of the business.
The willingness of CEO and the board etc to stick around during the turnaround process is a key risk. I think that the CEO could stay but the board may leave as they clearly wanted to sell the company.
Board and CEO alignment is low as I noted.
Tower is an insurer. If anything I think its claims expenses and risk management processes will tighten significantly via the introduction of things like address-based pricing as I mentioned. However, its risk pricing assumptions are really unknowable, and additionally the company does carry ugly worst-case scenario risks. There could be a monster storm that flattens New Zealand and bankrupts the company.
Tower’s Pacific island business also seems quite vulnerable in this regard. It would be a nasty surprise to find that the company has been pricing insurance wrong or not reinsuring appropriately, and there is no real way of evaluating that for a fact. We are in the La Niña part of the cycle which is associated with higher rainfall which is worth keeping in mind. (I also think that the Islander businesses are a decent hidden opportunity over the ultra long term, but I am not ascribing much value to them presently.)
Legal dispute with Peak Re, as mentioned above.
The bottom line:
With all of the opportunities for improvement – assuming the company is pricing risk correctly and reinsuring adequately – I simply think it is very unlikely that Tower’s price today is the right one. It may not end up a huge winner, but I think it is a good example of an asymmetric opportunity with a limited downside, and I have made it one of my larger positions.
Data on my purchases is contained in my original purchase piece, here.
I own shares in Tower Insurance. I have no financial interest in any other company mentioned. This is a disclosure and not a recommendation.
Comments: 2
Any thoughts on Bain Capital taking over Vero’s stake? They’ve even got an escalation clause in the bargain to pay up the difference to Vero if they mount a takeover bid. Might be good news for you!
Generally I think it’s good news although I also suspect it might lead to a takeover (I’ve sent some questions off to management on this and other topics, if they respond and agree I will publish them to my blog). I hope that whatever happens it is with a long-term turnaround in mind. I would prefer to see Tower not get taken over, although I think it is clearly undervalued and I would not be surprised to see someone else (e.g. Bain, or Fairfax) making another tilt at the company.