It has now been seven months since AmTrust decided to delist all six series of their preferred stock. Conversation about these securities, while active and impassioned for the first few months after the delisting, has more or less died down in the last several months. I believe the preferreds, currently trading in the range of $14 to $15, are on the verge of a significant move upward, but before we get to that, let’s try to understand what happened, what the market’s reaction was, what the arguments for and against suspending the dividends were and are, and see if we can think about what’s most likely to happen moving forward.
*Note - After this article was submitted to Seeking Alpha editors mid-day on August 7th, AmTrust declared the next preferred dividend after the market closed last night. That dividend validates and confirms the argument below.
AmTrust taken private
AmTrust officially completed their take-private transaction on Nov. 29th, with the E series (just to pick one of the series) closing that day at $14.60. The Karfunkel and Zyskind family teamed up with Stone Point Capital, as well as Enstar (NASDAQ:ESGR) and Madison Dearborn (more on Enstar later), to buy the approximately 45% of AmTrust not already under control by the family. At the time of the privatization, the consensus opinion regarding the preferred dividends was that they would be eliminated immediately after the transaction closed. Never mind that they had just announced the November preferred dividend (to be paid post-closing) – that was irrelevant according to the consensus opinion. They had announced it prior to the transaction closing, so the fact that it technically occurred while AmTrust was already private was apparently a moot point.
For anyone evaluating AmTrust preferred stock between late last year and this year, there were two potential changes that could happen to the shares:
- AmTrust could choose to delist the preferred shares.
- AmTrust could choose to suspend the dividend on the preferred shares.
But, if you think about it, the listed status is not central to the analysis. So long as the dividends continued along with general health of the balance sheet, whether the securities were listed or not was actually irrelevant to the investment decision. Surely anyone would participate in an investment in a private business offering a high enough return on capital, even though by definition it would be unlisted on a major stock exchange.
Further, while there were some people saying delisting was on the table, the conversation among investors was predominantly on the cessation of dividends and not on the delisting. Which was as it should have been. We have become over-dependent and used to a highly liquid investment environment, but the truth of the matter is that the listed status of an investment opportunity is largely immaterial to its merits.
The fireworks began after the market closed on Friday, Jan. 18th. AmTrust announced that they planned to delist the securities mentioned above with an effective date of Feb. 7. Whereas the E series closed Jan. 18th at $16.02 (12.09% yield), they closed the following Monday the 22nd at $9.85 (19.67% yield) – a 40% decline in one day. They hit a low of $7.80 on Jan. 31st, implying a yield of 24.84% if dividends were to continue – a 51% decline from the Jan. 18th close and now selling at 31% of par value. Investors (or speculators) were bailing in droves. The price collapse was likely significantly related to institutional investors puking their shares en masse - likely because of restrictions on owning unlisted securities
The delisting created a chaotic environment in the preferred securities, both in the trading price and in investor sentiment. Sentiment was overwhelmingly negative, largely influenced by the scores of investors that had bought at higher prices and were now sitting on large paper losses (or, for some, real losses as they bailed at January’s bottoms). Understandably, the conversation post-delisting centered immediately on the continuation of dividends. To anyone looking at AmTrust preferred securities in the days and weeks after the delisting, really the only objective of the analysis was to answer the following question:
Will the dividends continue to be paid, or will they be suspended?
That question is at the core of any analysis of AmTrust preferred securities since November, and the overwhelming answer from investors has been:
NO. Of course not! What are you, nuts?
Arguments for cutting dividend
Some of the reasons given for the dividend being suspended are not illogical. Preferred dividends are not mandatory, and preferred holders cannot force companies to pay them. They are purely at the discretion of the Board of Directors. Further, AmTrust preferred dividends are non-cumulative, so scrapping one (or ten) won’t lead to an unpaid dividend liability piling up for some later date.
Arguments like the following were advanced:
“Zyskind et al can (and IMHO likely will eventually) suspend the pref dividends. It would save them something like 68MM per year (from memory). There is no legal obligation to keep paying these (and they aren't cumulative), so why would they? Stopping payment (A) helps their equity position, b/c capital builds up at the holdco level and (B) helps their debt ratings (for the same reason as (A).”
“People keep saying, "We're fine, since the common shares can't pay divs if they suspend our pref divs." This is IRRELEVANT. Zyskind didn't take this private at a large premium in order to pay dividends to himself (he could have done so with it public). He is going to let operations recover for several years (or more) and then either IPO it again or sell it to the highest private bidder. And he will likely be very happy to suspend your dividends in the interim. Note that if he sells it to a PE firm in, say, three years, he will probably make a large profit on his common equity - yet the buyer doesn't have to pay the pref dividends either after the change of control! So the pref holders would STILL be screwed.”
“Does Amtrust have a legal obligation or a contractual obligation to maintain a registration for the preferred stock and to file financial statements with the SEC? No, they don't. Does Amtrust have a legal obligation or a contractual obligation to declare and pay quarterly dividends? No, they don't. The above thread talks about benefiting the common at the expense of the preferred by deregistering and saving a few million a year in fees. They may soon be benefiting the common at the expense of the preferred by not declaring preferred dividends to the tune of over $65M a year - that's what you should focus on. Where do you think the deregistered preferred trades after they fail to declare six dividends????”
“I doubt there is any rule that prohibits a strong company from suspending the preferred dividends. If you think their strength guarantees that they will pay or the fact that they are able to pay means a lot, I think you are deceiving yourself. Since they no longer pay common dividends, why pay the non-cumulative preferred dividends. You do it if you are ethical like Partnerre (PRE preferred stocks). If you are a Karfunkel, who knows.”
What’s curious to note is that the arguments given for why AmTrust can and should suspend preferred dividends – that they don’t have to, that it’s not an obligation, etc. – didn’t start with the privatization or with the delisting.
Put another way, the lack of any obligation to pay preferred dividends existed the day they issued the first series six years ago. Lost in this argument is that the public or private status of the firm doesn’t impact the economic logic of suspending preferred dividends. Why didn’t AmTrust suspend preferred dividends while still a public company? Why did AmTrust ever pay a single preferred dividend? Why didn’t they raise preferred equity and cut the very first dividend? Presumably, the response would relate to not wanting to cut off their access to future preferred equity raises, but certainly that still applies as a private business. If people want to argue using cold, economic, legal-based logic, then let's put that logic on steroids and go back to the first quarter after you received your preferred equity investment. Why even pay that first dividend? Could we be missing something?
All non-cumulative preferred stocks are under no obligation to pay dividends, but in some ways, that has to be too extreme, as clearly preferred stocks as a type of security could only exist if the expectation was for dividends to be paid, regardless of contractual mandates. Stiffing preferred holders has to be the exception rather than the norm, otherwise preferred equity wouldn’t even exist.
Further, the fact that other companies (and other private equity companies as well) have decided to not suspend preferred dividends after buying out the common forces the logic here into the corner of “well, these guys are shadier.” That may well be so, but it’s unquestionably a subjective opinion based off mediocre assumptions. “They said they wouldn’t delist and they did; therefore, that also means they’ll suspend the dividend.” That’s a weak argument.
Additionally, there is a very similar situation currently happening with the private equity firm Apollo and Aspen Insurance. Apollo bought out Aspen but left the preferred stock outstanding. They have since paid two dividends as a private company. So, both Aspen and Amtrust have been taken private in the last year by private equity firms, both decided to leave the preferred shares outstanding, and both have non-cumulative shares. Where are all the detractors talking about how Apollo will most certainly cut the $30 million in preferred dividends to AHL-C and AHL-D holders? Where’s all the talk about benefiting Apollo shareholders to the tune of $30 million a year at the expense of preferred holders? The arguments for AmTrust cutting the dividend seem exclusively related to intention, malice, and cold economics. What you don’t see are reasons due to their financial ability to pay the dividends. Shouldn’t the same cold logic apply to Aspen as well? The AHL-C and AHL-D threads on Seeking Alpha are ghost towns, without a single peep about how Apollo could easily suspend the dividend if they want to.
AmTrust securities trade around $14 whereas Aspen's two series are at $25.27 and $26.87. Are Aspen’s preferreds really worth 80-90% more? And yet the drumbeat about AmTrust suspending their dividend was loudest when the shares were at their nadir, trading around 30 cents on the dollar. This seems to be a misunderstanding of how risk and reward works, and an example of reflexivity in the market. As price declines - all else being equal - the risk/reward by definition gets more favorable. All of this would seem to suggest that a lot of the gloomy commentary hiding behind this veil of dispassionate logic is actually being influenced by subjective and emotional feelings about AmTrust’s management.
Arguments for paying
Let’s now turn to arguments for why preferred dividends were and are likely to continue post-delisting.
- Really the only argument for cutting the preferred dividends is so that the company can save money. There is no other reason. So, if we go down that road of wanting to save money, why would they choose to pay the dividend at the first opportunity to eliminate it after the acquisition? Lost in the delisting mayhem was that they simultaneously issued a second release announcing the March 15th dividend. This was now their second dividend as a private company and the first to be paid as a delisted company. What would be the reason to pay one dividend at a cost of $16 million only to then stop paying it at some arbitrary future date? Doesn’t the argument for saving money imply you should implement it as soon as possible?
When I posed this question, I got responses like the following (my comments in bold):
“To not have the price of their securities crash too much while they're still quoted? (Bad for public perception of the "brand"?) Right, because them delisting and cutting the dividend means they really care about the image that retail investors buying preferred securities at par have of them (also, as if doing just the delisting only to do the suspension later makes the perception: “well, at least they didn’t do it all at once.”)
To not attract too much press/scrutiny until the delisting has occurred? This makes no sense.
To leave time for some people they know to unload shares of these securities?” Probably illegal and seems to be flailing for any response to a very valid question.
“Probably they will wait until the shareholder uproar about the delisting dies down a bit.” Why would they care? Why would a company that plans to delist and suspend really care if it happens spaced out over a few months or all at once? And is avoiding a little bit of shareholder uproar really worth $16 million per quarter to them? Aren’t you kind of arguing against yourself at this point? If the argument to cut the dividend is purely because they don’t have to and they’re better off by doing it, aren’t you contradicting yourself by essentially saying that they’re going to burn $16 million a quarter ($50 million cumulatively thus far) to be delicate with shareholder feelings?
2. AmTrust has likely had a plan regarding the dividends for months, and if we go back before the acquisition closed, the possible options were as follows:
Option 1: Continue paying as if nothing changes.
Option 2. Stop paying immediately after going private.
Option 3. A hybrid wherein they pay some, then stop.
Option 4: A hybrid wherein they stop, then restart again.
When the first dividend after the acquisition was announced, options 2 and 4 were ruled out. That left options 1 and 3. While theoretically possible, option 3 doesn’t seem to make any sense. What rationale would the board come up with to pay some arbitrary number of dividends only to then stop. With three dividends being paid since the acquisition so far, under what logic would the company pay $50 million in dividends if they didn’t have to, only to then stop? The argument often given for stopping the dividend payments is one of saving money; why then would the board opt to pay some dividends ($50 million and counting) only to then start saving money? Under this light, option 1 appears far more likely than option 3, and it’s critical to understand that option 1 only gets that much more likely than option 3 with each passing dividend. In a sense, option 3 carries with it a contradiction. The longer the dividends go prior to stopping, the weaker the argument for stopping becomes (as the argument for stopping is to save money, but the longer you wait to stop the more money you've wasted).
3. In the press release announcing the delisting, what would be the reason to state the intention to continue paying quarterly dividends if that wasn’t at all tethered to reality? It doesn’t seem to make any sense why you’d say that only to then go directly counter to that in three months at the next dividend declaration date. The common catch-all argument here is that “these are shady people.” True. But having shady management doesn’t always lead to the worst possible outcome for shareholders at every decision juncture. Of course, they’ve lost the benefit of the doubt; it’s just not a stretch to question why they’d even bother stating the continuation of dividends when they really have no need to. Why volunteer completely inaccurate information that you were under no obligation to provide? Sure, they said they weren’t going to delist and then they did, 10 months later. Here, they were the day they announced the delisting, stating that, by the way, we plan to continue paying these. Although I admit this is potentially too granular, it is nonetheless interesting that in the press release announcing the delisting, they even put the fact that they intend to continue paying dividends in bold.
The common response to this is, “they’re probably lying.” Well, again, what incentive do they have to lie about something that they had no obligation to even mention? Two dividends later, what if they weren’t lying?
4. No one seems to be paying attention to the fact that Enstar, one of the three companies involved in the take-private, is publicly traded. Let’s go back to some of the arguments given for why the dividend would definitively be cut:
“People keep saying, "We're fine, since the common shares can't pay divs if they suspend our pref divs." This is IRRELEVANT. Zyskind didn't take this private at a large premium in order to pay dividends to himself (he could have done so with it public).
“Since they no longer pay common dividends, why pay the non-cumulative preferred dividends?”
Well, unfortunately for those two comments, Enstar’s publicly available documents make those two statements untrue. As Enstar states in their 10-K, they recorded dividend income from AmTrust of $0.3 million for their one month of AmTrust ownership in 2018. In their 10-Q from May 8th, they reported $1.8 million of dividend income from AmTrust for the first three months of 2019. And on Tuesday night, they reported another $1.8 million of dividend income from AmTrust for the three months ending June 30th. So, two straight quarters this year, two straight $1.8 million dividends. Given that they own 7.5% of AmTrust’s common shares, these are common dividends - meaning they cannot be paid unless the preferred dividends are paid. And the amounts seem like a fairly standard common dividend yield. $1.8 million per quarter would be $7.2 million per year, or a 3.6% annual dividend yield on their $200 million investment.
5. They have frequently raised capital through preferred stock (six series), with the most recent preferred issue being only three years ago. Stopping payment would seem to hurt their ability to raise preferred equity in the future. Additionally, and this was a point mentioned to me by Joe Eifrid, AmTrust sought out this preferred equity capital themselves (again, only three years ago for some of the series). Some of the delisting/suspension horror stories thrown around (like Equity Inns, which we’ll get to below) often involve new company X buying out old company Y, with new company X having no ties to the issuance of Y’s preferred securities. AmTrust management hasn’t really changed, and there isn’t a new majority owner; I think they will honor the agreement they themselves signed up for when issuing the preferred securities.
6. While I am not sure or entirely convinced that a dividend suspension would have a material impact (or any) on their A.M. Best rating, Best has said that access to capital markets is viewed positively in the ratings process – another reason not to remove preferred equity from your capital raising quiver.
7. There have been a number of companies that delisted their preferred stock after an acquisition that continued paying dividends as if nothing happened. Examples include:
- Bristol Myers-Squibb delisted its preferred stock in 2013. Dividends have continued, and it’s still outstanding.
- MB Financial delisted its non-cumulative preferred stock in 2019 and has now declared two dividends. Trades over par.
- Pitney Bowes delisted preferred stock in 2013. They continued paying dividends and just redeemed them in 2019 for $28.
- Central Maine Power Company delisted several preferred stocks in 2002, ultimately called them at par at various times. Dividends continued.
- El Paso Tennessee Pipeline Co delisted preferred in 2002, called at par in 2005. No sign that dividends stopped (hard to track with old delistings).
- Accredited Home Lenders delisted in 2007 before recession, went bankrupt, eventually liquidated and paid total liquidating distributions of $19.27 per preferred share.
- First BanCorp. voluntarily delisted in 2011 after offering to exchange for common. 89% of preferred holders converted to common. Has paid dividends on the ones that didn’t exchange since. Trades over par.
- Perini delisted in 2004, called in 2006 at par.
- American Land Lease got acquired, left preferred outstanding, delisted early 2009 during recession. Immediately suspended dividend. 5 quarters later it was reinstated, shares were called at par 4 years later.
- AIG bought Validus in 2018, left preferreds outstanding. Redeemed them three months later at par.
There is one messy example, often cited by detractors of AmTrust preferreds:
Equity Inns was acquired by Goldman in 2007. They paid 2-3 dividends, then stopped. Preferred holders sued and ultimately settled in late 2015 for $26 per share. Obviously, this is one of the messiest. There were several interesting facets here. First is that this was a real estate company acquired just before the recession in a very leveraged buyout, with the terms of the loans that Goldman took on limiting preferred dividend payments. Second is that the price stayed $20-$25 post-acquisition, while dividends continued, and only after suspension dropped to the range of $2-9. Anyone buying in 2008 at $9 walked away with a 16% annualized return when the settlement came through seven years later; anyone buying at $5 walked away with 27%. There are even people that picked up preferred shares around $0.24, implying a 100-bagger over seven years or a 91% annualized return.
What is also interesting is that some on SA have commented that Equity Inns is different, given that the preferred stock was cumulative. I think that’s a weak point. The reason is that actually by the time settlement talks began, cumulative unpaid dividends stood at more than $10, meaning that the actual amount due the preferred holders was over $35. Yet the cumulative unpaid dividends didn’t appear to factor into the settlement number. The $26 agreed upon seemed more based on the principal value of the preferred equity investment plus a dollar for the wait/annoyance.
General thoughts about this list before moving on to #8:
Two things seem clear. The first is that the preferred obligation is real. By that, I mean that eventually the preferred holders are redeemed or paid out some portion during a liquidation. Their stake cannot realistically be ignored perpetually as if it never has to be paid back, assuming the equity is there. I understand that in an academic sense, the shares truly are perpetual, but in reality, the preferred equity slot on the balance sheet does represent a real obligation that is redeemed at some point (if there is equity to cover it). Even in cases where dividends were halted for extended periods of time, ultimately, the company took out the preferreds and, short of a liquidation, the redemption took place at or above par. Perhaps the strongest proof of this point is seen in Goldman’s actions during the Equity Inns case. Even though they suspended the preferred dividends, they still spent $60-70 million buying back preferreds on the open market when the prices collapsed. You would only do this if you knew that, regardless of your dividend choices, the preferred obligation on your balance sheet still remained. Why spend good money reducing a balance sheet obligation that could be ignored perpetually?
The second is that delistings don’t appear to correlate strongly with dividend suspensions. Indeed, not in this analysis are the listed preferred stocks that suspended their dividends (Maiden (NASDAQ:MHLD), SunEdison (OTCPK:SUNEQ), Wheeler Real Estate Investment Trust (NASDAQ:WHLR), PG&E (NYSE:PCG), for example).
8. AmTrust mentioned in their defense to KBW that KBW hadn’t alleged that they’d suspended a dividend, nor had they. While this is likely standard legal writing, it would nonetheless be odd to then suspend the dividend, as that statement already used in one of their litigation defenses would be contradicted.
9. As referenced above, any statement alluding to AmTrust having no obligation to pay preferred dividends neglects that this has always been present – going private didn’t change the terms of the security. All preferred stocks are under no obligation to pay dividends. It’s interesting to note that I don’t see AmTrust preferred detractors posting on the Seeking Alpha pages for all the preferred stocks trading at par that the dividends could (and should) be suspended any moment. To focus exclusively on preferreds that were already trading at a 70% discount is, in my opinion, to be irrationally influenced by the share price and to become more pessimistic in a reflexive way based on how the market was reacting to the share price collapse. And again, it misses the point that the obligation never existed – going private didn’t change that.
10. Delisting does not signal an intention to suspend the dividend. They are two separate, unrelated events. Indeed, I think a lot of the people that believed the dividend would be cut (especially right after the delisting) were being irrationally influenced by their emotions about the delisting. One doesn’t imply the other, nor does one make the other more likely to occur.
Where do we go from here?
Thus far, there have been several groups proved correct. The first group consisted of people telling others to bail out of the preferreds early in the buyout process, with preferreds still near par. This group, while initially correct, doubled down later in 2018, warning of dividend suspension immediately following the take-private, hypothesizing that shares would reach the range of pennies per share. Alas, this group was proven wrong with the first dividend being announced simultaneous with the delisting announcement. Of course, the first and second dividend proved nothing conclusively, but any student of logic was left wondering: If cutting the dividend was as straightforward as many had concluded, why didn’t it happen? If you’re able to slash an expense, why didn’t you?
The answer to this is still unclear, but at the very least, it should invite the consideration that perhaps many people following the situation missed something in their mental calculations. Many arguments before the first dividend was announced carried bizarre assumptions of the Board being whimsical – likely another example of an improper extension of reasoning (“they lied about the listing status; therefore, they likely don’t have a plan and make it up as they go.”) Make no mistake, the fate of the preferred dividend was discussed during the take-private negotiations and they most certainly had a plan (and currently have a plan) for how they would handle it as a private, delisted company.
I think there are several things still being missed in the market’s analysis of AmTrust preferreds:
1. Among paying preferred securities, AmTrust's are second cheapest out of 412 different securities. That means they are currently cheaper than ~99.5% of all paying preferred stocks. To any value investor worth their salt, that obviously means they are at least worthy of analysis and, further, that the analysis is different than the average preferred selling at par. As price declines - all else being equal - the risk/reward by definition gets more favorable.
2. The analysis at this point doesn't have to stretch decades into the future (or quite possibly even a few years). If the dividends continue, the price will not stay $14 per share. That is definitive. Who on here is willing to argue that, were dividends to be declared in August and November, that we'd still be staring at $14 securities?
3. Continuing from point two, given that AmTrust has declared two dividends as a delisted company, what is the probability they pay a third? Forget about 20 more dividends, what's the probability of just getting to the third dividend? Easily over 50%, and I'd argue much higher. Now, assuming they pay the third dividend, what happens to the shares? What happens if they pay a fourth?
AmTrust has announced third quarter dividends the last two years on August 8th. With two dividends under their belt as a delisted company, with Enstar revealing this week that they got their second straight common dividend, and with no obvious and strong reason why dividends won’t continue, how likely is it that AmTrust declares their third straight dividend as a delisted company this week or next? Again, easily over 50%, and likely over 90%. What happens to the share price if that happens?
Lost in a lot of the conversation is that the analysis doesn't require owning these securities for 10 years (or even 10 months). Now, any investor should be prepared to, but the most likely scenario over merely the next few quarters (or merely the next couple weeks) is that dividends continue as usual, and if that happens the price will move dramatically from current levels.
Now, of course, I could certainly be wrong. I don’t know for sure what will happen, just like everyone else. But I think that’s a point that also is being missed in this analysis. Successful investing is not about finding certainties. As Marks teaches us, things that are supposed to happen fail to happen all the time. Things that shouldn’t work end up working all the time. To cope with this, we have to engage ourselves in the world of probabilistic thinking. Being wrong on a bet that you felt had a 90% probability of success was still the right decision, and you should take them every time. No one is certain how AmTrust preferreds will ultimately play out, but who can argue today that it is more likely than not they suspend the dividend this week or next? It seems clear, after three dividends as a private company, that the probability overwhelming points towards the continuation of dividends, and if that happens the share price should move dramatically.
Disclosure: I am/we are long AFSIM, AFSIC, AFFS. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.