AmTrust Financial Services: A P&C Blow-Up In The Making - Part I

| About: AmTrust Financial (AFSI)


AFSI exhibits all the traits of prior P&C insurance blow-ups.

Analysis of NYDFS conditions imposed on AFSI indicates regulators may share the shorts' concerns.

Questions raised by critics about AFSI's accounting remain unanswered.

Price target: $0 (Insolvent).


I believe the equity in AmTrust Financial Services, Inc. (NASDAQ:AFSI) is worthless, and I believe the market will soon come around to my view as a number of catalysts fall into place to expose what I believe are substantial accounting violations at AFSI.

One of these catalysts is increased scrutiny of AFSI's accounting and financial condition from regulators, particularly the New York Department of Financial Services ("NYDFS"). I believe additional oversight from the NYDFS will put direct and indirect pressure on AFSI to improve internal controls, use more reasonable accounting assumptions, and correct material misstatements.

Given how leveraged AFSI already is on an unadjusted basis (see below), I believe even a small accounting correction would be catastrophic for the company and the stock. However, the magnitude of the Company's misstatements is anything but small. Therefore, I expect AFSI to join the list of once-high-flying property and casualty carriers who inevitably crash and burn when they are no longer able to outrun their reserving and other problems with additional growth.

This article is organized as follows:




Business Overview


Similarities to Prior Property & Casualty Insurance Company Failures


Analysis of NYDFS Conditions Validates AFSI Critics


Summary of Potential Accounting Violations at AFSI



I. Business Overview

AFSI's chairman, Michael Karfunkel, along with his brother George Karfunkel (a director) and his son-in-law Barry Zyskind, the CEO, founded AFSI to purchase a computer warranty business in 1998. Since then, the company has grown from $10 million of gross premiums written to over $5 billion of gross premiums, through a combination of acquisitions, expanded geographies and/or products

Of AFSI's $3.1 billion of gross premiums written in the first half of 2014, approximately 44% were associated with workers' compensation policies the Company issued. Warranty is the next largest line of business at 16% of year-to-date gross written premiums, followed by other liability policies which represented 12%, and commercial auto at 7% of AFSI's premiums.

AFSI reports results in three segments (a fourth segment is in runoff):

  1. Small Commercial Business, the vast majority of which is workers' comp;
  2. Specialty Risk & Extended Warranty, which houses all of AFSI's non-U.S. business, along with its U.S. Warranty business; and
  3. Specialty Program, which writes a combination of commercial auto, other liability, and some workers' compensation.

Across the Small Commercial and Specialty Program segments, AFSI was the 8th-largest workers' comp underwriter in the country in 2013.

Additionally, AFSI's significant quota share partner, Maiden Holdings (NASDAQ:MHLD), is a related party due to its shared controlling shareholders, the Karfunkels.

II. Similarities to Other Property & Casualty Insurance Company Failures

The history of property and casualty insurers is littered with previously high-flying companies that thought they had built a "better mouse trap" (or at least told investors they had) that allowed them to generate superior returns. With their supposedly "better mouse trap", these upstarts grew rapidly and investors crowed about their brilliance... for a while.

However, the story of a rapidly growing, casualty-focused insurer always seems to end the same: badly. I expect AFSI to end just as badly, and I expect it to happen soon.

Of course, high-flying insurance companies are almost oxymoronic, because the primary way an insurance company outperforms over long periods of time is by avoiding the temptation to grow too quickly, since they are subject to such severe adverse selection. That is why the high-flying new entrants inevitably fizzle out (or crash and burn).

As discussed in S&P's report "What May Cause Insurance Companies to Fail", AM Best's Annual Impairment Studies, and the Federal Insurance Office's report "How to Modernize and Improve the System of Insurance Regulation in the United States", these flame-outs and failures typically share at least a few of the following attributes:


Historical Example(s)

Relevance to AFSI

1) Rapid Premium Growth

Reliance & Tower

3-year CAGR: 48%

2) Frequent Acquisitions

HIH & Tower

10 acquisitions closed since 1/1/2013

3) Unrealistic Pricing & Reserving

Independent & Tower

Significant, persistent gross adverse development

4) Aggressive Use of Reinsurance

Independent & HIH

Maiden Quota Share & Luxembourg Captives

5) Unnecessarily Complex Structure

HIH & Reliance

Numerous subsidiaries & frequent changes

6) Heavy Reliance on Debt Funding


Significant debt outstanding & utilizes repo market

7) Excessive CEO Pay


$27 million in the last three years

8) Poor Governance Structure

Independent & HIH

Board's Executive Committee members are all related

9) Weak Financial Controls

Independent & Tower

Earnings Releases ≠ Quarterly Filings

10) California Workers' Comp

HIH & Tower

2011 entry followed by rapid growth

For additional background on previous insurance failures listed above, please see the links below:

  1. Independent Insurance
  2. HIH Insurance
  3. Reliance Group Holdings
  4. Tower Group International

III. Regulators Appear to be Catching On to AFSI's Shenanigans (Thus Validating The Shorts)

While AFSI has received a fair amount of press over the last year (most of it negative), it was only recently that regulators began to take action, at least publicly. Specifically, the New York Department of Financial Services' ("NYDFS") approval of ACP Re's acquisition of Tower Group International, Ltd. was subject to five disclosed conditions (per AFSI's 8-K), which are listed and discussed below.

Before analyzing the actual conditions, I think it's important to note that imposing conditions such as these is unprecedented, as far as I can tell. Not only that, they were imposed upon an affiliate of the acquirer (ACP Re), which may indicate just how concerned the department is about AFSI's condition and accounting.

A natural question to ask is: If AFSI was not happy with the terms, why would it agree?

I would argue that AFSI had no choice but to accept the conditions, because if AFSI balked at the prospect of providing the NYDFS with information or changing auditors, then the NYDFS would have KNOWN they had to investigate ASAP. However, by accepting the terms, AFSI at least bought itself some time.

Let's look at each of the conditions individually.

Condition #1: Strengthen actuarial resources

Form 8-K Excerpt:

"In accordance with an actuarial plan (the "Plan") submitted to and approved by the Department, AmTrust will strengthen its actuarial resources to ensure by year-end 2015 that its actuarial scale and capabilities are commensurate with its size. This will entail both hiring more actuaries, and ensuring that those currently on staff are adequately credentialed. Further, in accordance with the Plan, AmTrust will better institutionalize the role of its internal actuaries by creating a Global Chief Actuary position, as well as Global Chief Reserving and Pricing Actuaries, all of whom will be integrated into AmTrust's executive management team."

My comments:

While seemingly innocuous (Sell-side interpretation: "hire some extra actuaries and give them fancy titles"), I think this condition is hard to understate. Insurance companies live and die by their reserves, and the fact that the NYDFS seems to be uncomfortable with the veracity of AFSI's reserves is incredibly concerning.

Two other relevant sets of data points color this condition further. First, Barron's raised concerns that AFSI's reserves in their article "Balance Sheet Risk Makes AmTrust Shares Vulnerable", by Bill Alpert (emphasis added):

"Earnings surely look good. In the 12 months through March, the company reported that its net income totaled some $3.90 a share-making shares of the business appear fairly priced at 11 times earnings. Those earnings, however, come mainly from insuring property and casualty, which is known to have a "long tail." That is, claims can be reported or litigated years after they arise, confounding what seemed like a profitable underwriting record. When that happens, profits can prove to be figments of aggressive premium writing and under-reserving for loss.

Is that the case with AmTrust? Questions about whether the company is under-reserved arise because of some puzzling accounting disparities: AmTrust and Maiden Holdings show a $400 million difference in their accounting for the same reinsurance activities; AmTrust's numbers for acquired reserves differ by $50 million in different parts of its financial reports; while the latest 10-K's tabulation of loss reserves leaves AmTrust with negative reserves for some years-an impossible accounting that would mean that policyholders would actually pay AmTrust millions for claims in those periods.


"Insurance investors usually study a company's annual tabulation of loss reserves for each year. But, as noted, AmTrust's 10-K indicates that its reserves remaining after claim payments for all the years up through 2007 are $95 million more than the amount for those years plus 2008 - which if correct, would indicate that policyholders will pay AmTrust that $95 million on the 2008 incidents.

Malone says the fault is in our analysis, not the 10-K. But the same analysis was used in an April industry study by AmTrust investment banker FBR Capital Markets.

Loose accounting would only add to concerns about AmTrust's underwriting rigor, given the evidence that the insurer picks unusually low estimates for its eventual losses. Compared with peers in workers' compensation insurance, where AmTrust gets half its revenue, the company ended up paying out a higher percentage of its original estimate for losses on accidents in the years 2006 through 2012, according to an analysis of the industry's "paid-to-incurred" ratios in FBR's April report."

Just in case AFSI's responses to questions not concerning enough, I would suggest readers call the Maine Bureaus of Insurance and request a copy of the transcript of the hearing to evaluate ACP Re's application for approval of the acquisition of Tower Group International and its subsidiaries.

In response to a question about AFSI's actuarial resources at the hearing, AFSI's CFO responded that the lead internal actuary is NOT a "full fellow" (page 49 of the transcript), but noted the use of an outside consultant, Charlie Gruber of SG Risk, who is a "full fellow".

Unfortunately, AFSI's choice of SG Risk raises more questions than it provides answers. In 2010, SG Risk (among others) was sued by investors, after it came out that the trustees of SG Risk's client, CRM Holdings, had been over 50% under-reserved in 2006. The lawsuit also discussed the December 2009 "Notice of Imminent Enforcement Action" that SG Risk's client received from the New York Attorney General, which stated:

"CRM and the Principals were assisted in implementing these improper practices by CRM's actuary, Charles Gruber, of the actuarial firm of SG Risk, LLC, and its accountant, Donald Neubecker, of the accounting firm of UHY LLP, who produced misleading financial and actuarial reports for the trusts. These practices had the net effect of artificially reducing liabilities as reported in the trusts' financial statements that CRM and the Principals filed with the WCB."

If SG Risk is employing similar practices on behalf of AFSI, the Company could be severely deficient in reserves. In light of AFSI's significant overlap with Tower Group's exposures and Tower Group's massive reserve charges, this would not be particularly surprising.

Given SG Risk's reputation, it would make sense for the NYDFS to commission an independent reserve analysis like the one CRM's client did to ensure that AFSI is not in a similar predicament.

Condition #2: Maximum NPW-to-Surplus of 3:1

Form 8-K Excerpt:

"For the year ending December 31, 2015 and thereafter, AmTrust will cause its insurance company subsidiaries to maintain, in the aggregate, a ratio of net written premium to surplus of 3 to 1. Net written premium shall be net of reinsurance, including business retroceded by AmTrust International Insurance, Ltd. ("AII") to Maiden Insurance Company, Ltd., a Bermuda insurer and an affiliate of the Applicants [ACP and affiliates], AmTrust and National General Holdings Corp. In support of the foregoing, AmTrust shall, subject to regulatory approval, modify its current intercompany reinsurance structure by reducing the quota share cession to AII to 60% effective January 1, 2015 and 50% effective January 1, 2016 and thereafter."

My comments:

AFSI discloses the statutory surplus of individual subsidiaries, but it does not disclose its consolidated statutory surplus, and frequently discusses its "total capitalization" under GAAP, which analysts sometimes confuse with statutory capital.

Due to AFSI's highly complex and constantly changing organizational structure, the Company's lack of disclosure with respect to consolidated statutory capital makes evaluating AFSI's compliance with this condition difficult.

However, the differences between GAAP and statutory accounting are relatively straightforward. The table below sets forth my estimate of consolidated statutory capital (Unadjusted: $502.7 million, Adjusted: $856.8 million).

As a result of this analysis, I believe AFSI needs to raise at least $292 million. However, this estimate is before considering the effects of the accounting concerns expressed in the next section of growth.

Incorporating 25% growth in the second half of 2014 (probably conservative), I estimate AFSI needs to raise over $400 million, only a small portion of which can be debt-funded due to leverage ratio covenants. Should my and others' accounting concerns be confirmed through NYDFS oversight or otherwise, the hole could be much, much larger.

Condition #3: Reduction in line 4 of AII's 2013 statutory balance sheet

Form 8-K Excerpt:

AmTrust will cause the intercompany receivables included on line 4 of AII's 2013 statutory balance sheet to be paid down on or before December 31, 2015, with at least 50% of the amount to be paid on or before December 31, 2014. The method(s) by which these receivables are paid off shall be subject to the review and prior approval of the Department.

My comments:

While AII's statutory balance sheet is not publicly available, the Bermuda Insurance Accounts Regulations 1980 indicates that line 4 represents investments in and advances to affiliates. In other words, this line represents AII's ownership in other AFSI subsidiaries.

Interestingly, this seems to touch upon an exchange between Mr. Alpert and AFSI (per AmTrust's response to Barron's):

Question 7: AmTrust subsidiaries' 2013 Schedule Y, Part 1A show AII owning about $300 million worth of affiliates like Rochdale, AIUL (Ireland) and AEL (NASDAQ:UK) that all retain some risk, so how accessible is that piece of AII's capital(?)

Company's Response to Question 7: Schedule Y part 1 does not show any capital/surplus amounts. Therefore, Schedule Y part 1 does not show that AII owns about $300 million worth of affiliates.

AII owns Amtrust International Management, which owns 50% of Amtrust Equity Solutions, which owns Rochdale, AIUL as well as IGI - AII's total indirect ownership of those three companies is $237 million.

The capital of Rochdale, AIUL and AEL is very accessible. The total statutory capital of the companies mentioned is $475 million and is 3.7 times those same entities required capital of $128 million. The company could reallocate capital if necessary. In addition, the combined liquid assets of Rochdale, AIUL and AEL is over $1.0 billion.

Editorial Note: While the Schedule Y, Part 1 does not show capital amounts, it does provide the ownership structure, which can be used in conjunction with Footnote 22 of AFSI's 10-K to get the numbers used in this discussion.

This is important because AII is already very thinly capitalized, as Barron's pointed out (see below), and AFSI does not have capital at the holding company that it can contribute to replace the assets that the NYDFS is requiring AFSI remove from AII.

The cantilevering of AmTrust premiums over its capital base is notable, because that base contains an element that differs from most American property and casualty insurers. AmTrust's local insurance units cede over half of their premiums and losses to the company's wholly owned "captive" reinsurer in Bermuda (which, as noted, then cedes risk to Maiden).

State insurance filings of AmTrust units show that the Bermuda captive has lost about $400 million under its reinsurance agreements with its AmTrust counterparts in the last five years. The Bermuda unit's regulatory capital fell last year from $499 million to $416 million, a level just over two-times the minimum required for "solvency" under Bermuda's relatively lenient standards. By contrast, Maiden ended the year with more than four times Bermuda's capital requirement.

AmTrust's Bermuda unit is more than adequately capitalized, says Malone, and its numbers shouldn't be compared with those of any other reinsurer. That's because it reinsures only stable, predictable risks, she says.

Condition #4: AFSI is required to change auditors

Form 8-K Excerpt:

In light of AmTrust's growth and increased geographic footprint, AmTrust will engage an external auditing firm with corresponding global resources and skills beginning with the audit for the annual period ending December 31, 2015. Before the engagement is undertaken, the selection of the auditing firm shall be subject to the review and prior approval of the Department.

My comments:

Even the most ardent AFSI bull must have been a little fazed by this condition, because (1) AFSI wasn't even the actual acquirer, and (2) it indicates DFS may be less than confident in AFSI's financial statements. In contrast, bears (myself included) seem excited about the prospect of a new auditor reviewing AFSI's financial statements.

I would add that BDO might decide to tighten their review so as to avoid the embarrassment of having a new auditor force a restatement of financials that were blessed by BDO.

Condition #5: Provide the NYDFS any and all financial information requested

Form 8-K Excerpt:

AmTrust shall, beginning with the month ending September 30, 2014, and for a period of not less than three years, provide to the Department any and all financial information as the Department may request.

My comments:

Yet again, the second Barron's article offers investors a clue as to what the NYDFS saw that prompted them to request access to any and all financial information:

"Insurance investors usually study a company's annual tabulation of loss reserves for each year. But, as noted, AmTrust's 10-K indicates that its reserves remaining after claim payments for all the years up through 2007 are $95 million more than the amount for those years plus 2008 - which if correct, would indicate that policyholders will pay AmTrust that $95 million on the 2008 incidents.

Malone says the fault is in our analysis, not the 10-K. But the same analysis was used in an April industry study by AmTrust investment banker FBR Capital Markets.

Loose accounting would only add to concerns about AmTrust's underwriting rigor, given the evidence that the insurer picks unusually low estimates for its eventual losses. Compared with peers in workers' compensation insurance, where AmTrust gets half its revenue, the company ended up paying out a higher percentage of its original estimate for losses on accidents in the years 2006 through 2012, according to an analysis of the industry's "paid-to-incurred" ratios in FBR's April report."

Furthermore, this request implies that the NYDFS did not have access to all financial information. Therefore, the suggestion that NYDFS approval was somehow a stamp of approval (as some sell-side analysts have suggested) appears to be unfounded.

Finally, I would note that Ocwen (NYSE:OCN) agreed to conditions with the NYDFS in conjunction with an acquisition. Anyone familiar with how Lawsky and the NYDFS have dealt with Ocwen know it was anything but a stamp of approval.

IV. Potential Accounting Violations at AFSI

Barron's, Off Wall Street, and GeoInvesting have all publicly criticized or raised questions about AFSI's accounting, and while AFSI's management claims to have silenced critics, I believe their responses raise more questions than answers.

Specific areas of concern raised by AFSI's critics include:

  1. Consolidation of Loss & loss adjustment expenses ("LAE") ceded to wholly-owned Luxembourg Reinsurance Captives ("LRCs");
  2. Discrepancies between the financial statements of AFSI & Related Party Maiden Holdings ("MHLD");
  3. Accounting for deferred acquisition costs;
  4. Discrepancies among AFSI's disclosures with respect to Loss & LAE Reserves assumed in conjunction with acquisitions; and
  5. Valuation of life settlement contracts, among others.

In the interest of brevity, I will only discuss the Luxembourg issue in this article.

AFSI's Explanations of Luxembourg Accounting: Inadequate and Different Depending on the Audience

AFSI shares fell precipitously last December after GeoInvesting wrote "AmTrust Financial Services: House of Cards?" In a hastily arranged conference call on December 16, 2013, management denied the allegation, but provided no answer beyond a flat denial and generalizations about Luxembourg captives, without answering the question of why AFSI seemed to ignore losses ceded to Luxembourg but did reflect the benefit to AFSI's Bermuda subsidiary

When Barron's questioned AFSI about why it did not reflect the losses of wholly-owned subsidiaries in its SEC filings, management gave a rather curious response (that the losses are excluded because they are inflated "artificially"):

"While the profit benefit appears in AmTrust's consolidated financials, losses it ceded to its Luxembourg units don't. AmTrust tells Barron's that those losses are inflated "artificially" or "synthetically" and don't reflect the real world claims on its primary insurance subsidiaries. "It's a way to draw down these reserves," says CEO Zyskind. "They are self-created losses within our organization, so they get completely eliminated in consolidation."

AmTrust says the arrangement is proper under U.S. and Luxembourg rules. That's interesting, since the company is reporting loss numbers to auditors and insurance commissioners that it acknowledges aren't authentic. The statutory filings AmTrust subsidiaries submitted to their state regulators over that same 2010 to 2012 period show nearly a quarter of a billion dollars-worth of these inflated losses going to Luxembourg. The synthetic losses also appear on the financial statements of AmTrust's Bermuda captive, AmTrust International Insurance, which it says are done according to U.S. Generally Accepted Accounting Principles under the eyes of the audit firm BDO USA."

Apparently, that comment did not go over particularly well with someone, because just days after the article was published, AFSI management stated that the loss are, in fact, real:

Robert Farnam - KBW

Hi there. Thanks. A couple of questions. It sounds like lot of the people that had with the short[ph] had issues with Schedule Y. So can you, I don't know Barry or Ron, can you just go over again how Schedule Y does work or doesn't work as it relates to your relationships with Luxembourg?

Michael Saxon

Schedule Y is a representation of transaction among companies within the holding company system, not limited to insurers transactions. To the extent that there is management fees or other organization vertically integrated organizations are disclosed in Schedule Y. Schedule Y on a particular column that people like to reference, that is a net number of premiums lost it. It's not like Schedule F in the reinsurance activity, it's just a recast among affiliated companies with the holding company. And again as Barry said in his comments, Schedule Y represents the net ceded loss to our Luxembourg captives based on our reinsurance structure.

AFSI's CEO also made contradictory remarks on the Q4'13 earnings call. First, he described the "stop-loss" agreement between AII and Socare:

"... We entered into a stop loss agreement between our Bermuda reinsurance company, AmTrust International and Socare. The agreement provides that coverage is triggered if AmTrust losses exceed certain attachment points, obligating the captive to pay Bermuda a maximum of $100 million in losses. In 2013 these stop loss agreements generated losses net of premium of $58 million to Socare. This results in the drawdown of $58 million of redundant equalization reserves and allows AmTrust to recognize a gain of approximately $5.2 million."

Socare was re-named AmTrust Re Theta on December 20, 2012.

One can retrieve the financial statements for AFSI's Luxembourg subsidiaries at and confirm that AFSI did cede losses in excess of premiums of approximately $58 million to AmTrust Re Theta.

Here's where the story breaks down. On the same Q4'13 earnings call, the CEO also said (emphasis added):

"... I want to address the references to Schedule Y in the annual statements filed by our U.S. insurers. Schedule Y is intended to show the net income effect of reinsurance agreements with affiliated insurers within a holding company group. Schedule Y shows correctly the aggregate net income effect to our Luxembourg companies resulting from their assumption of intra-company financial losses for AmTrust International."

However, the Schedule Y of AFSI's US subsidiaries tells a different story:

Once again, AFSI's "explanation" falls apart under scrutiny. Perhaps one of the cheerleaders universally buy-rated analysts will ask the CEO to explain these contradictions on the earnings call next week.

VI. Conclusion

As discussed above, AFSI has all the features of a high-flying insurance company destined for disaster. The second critical element, regulatory scrutiny, is also now in place. Finally, a review of AFSI's attempts to explain away its accounting criticisms indicates that the reason it cannot provide a clear answer is that the critics are on to something.

Considering AFSI's tangible book value is only $888 million, and losses ceded to Luxembourg, which do not appear to be reflected in AFSI's SEC filings, exceed $250 million, and AFSI's discrepancies with MHLD are nearly $400 million (per Barron's), it is easy to imagine a scenario in which the remaining tangible book value gets wiped out, just as it did at Tower Group, when it finally admitted to being severely under-reserved.

As such, my price target is $0. In future articles, I will provide additional details for why I believe AmTrust is insolvent and what catalysts I expect to precipitate AFSI's final descent.


In preparing this report, I have relied solely upon publicly available documents, such as AmTrust filings with the Securities and Exchange Commission ("SEC") and similar agencies, foreign and domestic, whose materials are available online. All sources used for the information set forth in this article are referenced herein. I have not obtained any nonpublic information, material or otherwise. The conclusions expressed in this report reflect my personal opinions, which are based upon the materials referenced herein. Please be advised that as a result of my analysis and the concerns expressed herein, accounts that I manage maintain a short position in AmTrust common stock and owns put options on the same.

Disclosure: The author is short AFSI.

The author wrote this article themselves, and it expresses their own opinions. The author is not receiving compensation for it. The author has no business relationship with any company whose stock is mentioned in this article.

About this article:

Tagged: , , , Property & Casualty Insurance,
Want to share your opinion on this article? Add a comment.
Disagree with this article? .
To report a factual error in this article, click here