Engine Capital Sends Letter to the Board of Directors of Houghton Mifflin Regarding its Intention to Not Tender its Shares into the Veritas Offer

Believes the Price Is Inadequate and Doesn’t Reflect the Value of the Company Following a Flawed Sale Process

Outlines a Superior Plan That Would Keep Houghton Mifflin Public Following a Dutch Tender Offer Between $21 and $22 Per Share

Announces Intention to Nominate Directors at the Earliest Possible Opportunity to Ensure Company Executes on Suggested Plan

NEW YORK--(BUSINESS WIRE)-- Engine Capital LP today announced that it has sent the below letter to Houghton Mifflin Harcourt Company's (Nasdaq: HMHC) Board of Directors.

***

Dear Members of the Board:

Engine Capital LP (together with its affiliates, “Engine” or “we”) is a long-term shareholder of Houghton Mifflin Harcourt Company (“Houghton Mifflin” or the “Company"), with ownership of approximately 2.6% of the Company’s outstanding shares, making us one of the largest fundamental shareholders of the Company. We are writing to let you know that Engine does not intend to tender its shares into the $21 per share offer from Veritas Capital (“Veritas”) for three fundamental reasons:

1. We believe the sale process and its timing were flawed.

2. We believe the $21 per share offer significantly undervalues the Company.

3. We believe there is a superior alternative available that would result in a higher per share value for all shareholders. This alternative plan would involve executing a large Dutch tender offer between $21 and $22 per share and keeping Houghton Mifflin public under the current management team. This would be a superior outcome for shareholders who want liquidity now and those who believe the Company’s stock is undervalued and want to continue owning the stock.

We don’t oppose this transaction lightly and you should know that since Engine’s inception almost nine years ago, we have never opposed an announced transaction involving one of our portfolio companies. In this particular case, however, we feel compelled to publicly do so because of the egregiousness of this transaction.

The Sale Process and Its Timing Were Flawed

We believe the process was fundamentally flawed. While it does appear at first glance that the Board reached out to 60 parties, the reality is that 58 of those 60 parties only saw the initial projections that were way too low.¹ Those 58 parties who decided to exit the process early and not bid for the Company did that after seeing the “initial forecast,” which included an expected 2021 adjusted EBITDA of $196 million (followed by an extraordinary jump to $270 million in 2022). However, 2021 adjusted EBITDA turned out to be 38% higher at $270 million (followed by a much more credible jump to $286 million in 2022), at which point the Company prepared a “revised forecast."² Unfortunately, those 58 parties were no longer involved in the process by then and never saw these significantly higher and more credible projections. Given the significant differences between the initial and revised forecasts, it’s as if you had run a sale process with only two parties – Veritas and Party B. Only those two parties saw the updated 2021 financials with an adjusted EBITDA of $270 million, an updated 2022 budget and an updated three-year plan. The 58 other parties only saw the “initial forecast” with an adjusted EBITDA of $196 million.

The $196 million of 2021 adjusted EBITDA in the “initial forecast” is also incomprehensible. This forecast was prepared in November 2021, when the Company already knew Q3 2021 results and at that point, year-to-date adjusted EBITDA was $246 million. The “initial forecast” for 2021 therefore assumed that Q4 2021 adjusted EBITDA would be negative $50 million when Q4 2020 adjusted EBITDA was a positive $16 million. Q4 2021 adjusted EBITDA turned out to be $24 million, a delta of $74 million versus the “initial forecast of negative $50 million, a difference that is hard to comprehend considering these projections were made in November 2021, in the middle of the fourth quarter. We believe this incomprehensibly low “initial forecast” played a large role in reducing the field of interested buyers from 60 to two and undermined the competitiveness of the sale process. The Company demonstrated a declining year-over-year Q4 adjusted EBITDA from $16 million in 2020 to negative $50 million in 2021, followed by a hockey stick adjusted EBITDA increase from $196 million in 2021 to $270 million in 2022. We suspect potential bidders found these numbers to lack credibility and as a result, exited the process, preventing them from seeing the much higher and more credible “revised forecast.”

We also find the timing of the sale process puzzling, with 2021 billings still below the low-end of the billings range provided by management at its 2019 Analyst Day. Based on the background section of the Schedule 14D-9, the Company initiated a sale process around August 2021. Why initiate a sale process at a time when billings were still below the low-end of the billings range? As shareholders, we were looking forward to the Company’s refinancing, the continued strong execution under the leadership of Jack Lynch, a normalization of billings and accretive growth from tuck-in acquisitions. We believe shareholders would have been better off if the Board had simply waited a couple more years until billings normalized. At that point, only two years from now, unlevered free cash flow per share would have been around $2.70, the stock price would have been significantly higher and the Company would have been in a better position to start a sale process.

The fairness opinion is also riddled with mistakes or imprecisions to try to justify the low $21 per share transaction. Just to cite a few, Evercore Group L.L.C. (“Evercore”), the Company’s financial advisor, used the Company’s net debt as of 3/31/2022 to derive different equity values. We question why Evercore would use the balance sheet at the end of Q1, when the cash position is at a low point and does not give any credit for the seasonally strong cash flow generation during the second half of the year. For context, in Q3 2021, the Company generated free cash flow of $265 million, the equivalent of $2 per share!³ In its public company trading analysis, Evercore did not consider the much higher multiples of education technology peers when considering the adjusted EBITDA metric. In that same public company trading analysis, Evercore failed to include the tax attributes valuation of around $2 per share when deriving the equity value of the Company. Finally, for its discounted cash flow analysis, Evercore used absurdly high discount rates ranging from 11.50% to 13.50%, which is way higher than what public market investors use in this low interest rate environment.⁴

We also note the significant conflict of interest by Evercore, which ran the sale process and rendered the fairness opinion. In particular, the Schedule 14D-9 states that: “In addition, during the two-year period prior to the date of its opinion, Evercore and its affiliates have advised / provided financial advisory or other services to Veritas for which Evercore received fees, in the aggregate, of approximately $75 million.”⁵

The $21 Per Share Offer Significantly Undervalues the Company

Over the last two years, management has repeatedly told us that the billings range it outlined during its 2019 Analyst Day remains valid (adjusted for the sale of HMH Books & Media). The table below highlights the unleveraged free cash flow of Houghton Mifflin if one assumes this billings range, the $850 million of fixed costs, and 65% flow-through above those fixed costs that management outlined on its Q3 2021 investor call.

	Low	Midpoint	High
Billings (including HMH Books & Media)	1,400	1,575	1,750
Less HMH Books & Media billings	195	195	195
Pro-Forma Billings	1,205	1,377	1,555
Fixed Costs	850	850	850
Variable Costs	124	184	247
Unleveraged Free Cash Flow (UFCF)	231	343	458
# Shares Outstanding	130	130	130
UFCF per share	1.78	2.64	3.53
Transaction multiple @21/share	11.8	8.0	6.0

At $21 per share, we believe Veritas is attempting to take Houghton Mifflin at 8.0x mid-cycle unlevered free cash flow. This multiple is simply too low for a business of Houghton Mifflin’s quality and with its growth prospects. The multiple is even more egregious when one considers that Houghton Mifflin is transitioning to a SaaS business model with an increasing percentage of recurring revenue and digitally connected billings, which is reducing the Company’s cyclicality. This means that over time, the Company’s multiple is likely to rerate toward the multiple of education technology companies, which is significantly higher than the multiple of education publishers. It is also worth noting that the Company has billions in federal and state tax loss carryforwards as of 12/31/2021, implying the Company won’t be a cash taxpayer for many years.

The projections filed as part of the Company’s Schedule 14D-9 confirm the bright prospects of Houghton Mifflin and are attached in the below table.

Revised Forecast and Extrapolations⁶

		2021A				2022E				2023E				2024E				2025E				2026E				2026E Norm. (1)
Billings		$	1,109			$	1,188			$	1,257			$	1,377			$	1,428			$	1,560			$	1,496
% Growth							7.1	%			5.8	%			9.6	%			3.7	%			9.2	%
Revenue		$	1,050			$	1,121			$	1,214			$	1,261			$	1,347			$	1,471			$	1,411
% Growth							6.8	%			8.2	%			3.9	%			6.8	%			9.2	%
Adjusted EBITDA		$	270			$	286			$	350			$	361			$	410			$	465			$	431
% Margin			25.7	%			25.5	%			28.8	%			28.6	%			30.5	%			31.7	%			30.5	%
Adjusted EBITDA Less Plate CapEx		$	214			$	229			$	292			$	302			$	353			$	407			$	374
% Margin			20.4	%			20.4	%			24.1	%			24.0	%			26.2	%			27.7	%			26.5	%
Adjusted EBITDA Less Total CapEx		$	175			$	180			$	244			$	255			$	304			$	354			$	324
% Margin			16.7	%			16.1	%			20.1	%			20.2	%			22.6	%			24.1	%			22.9	%
(1) 2026E Normalized is adjusted to account for the multi-year cyclicality in core curriculum purchasing cycles in California, Florida, and Texas (“Big 3”) and is based on the 8-year average of Billings between 2017 and 2024E to account for that cycle. Big 3 financials assume the same Big 3 Adjusted Cash EBITDA% margin and CapEx % of Billings as 2026E.

Based on the Company’s own billings forecasts, we can derive the unleveraged free cash flow per share to 2026 which further highlights the inappropriate multiple paid by Veritas for the Company.

	2022E	2023E	2024E	2025E	2026E	2026E Normalized
Billings	1,188	1,257	1,377	1,428	1,560	1,496
Fixed Costs	850	850	850	850	850	850
Variable Costs	118	142	184	202	249	226
Unleveraged Free Cash Flow (UFCF)	220	265	343	376	462	420
# Shares Outstanding	130	130	130	130	130	130
UFCF per share	1.69	2.04	2.64	2.89	3.55	3.23
Transaction multiple @21/share	12.4	10.3	8.0	7.3	5.9	6.5

The $21 per share offer is inadequate if Houghton Mifflin were to remain a standalone entity, but that offer is even more inadequate considering that Veritas will – in all likelihood – eventually combine Houghton Mifflin with Cambium Learning, a complementary asset that Veritas owns. We believe there are significant revenue and cost synergies between these two businesses that will exclusively accrue to Veritas. For example, both companies have large sales forces targeting school districts that could be combined and rightsized. There would also be significant cross-selling opportunities by attaching Cambium Learning’s products to Houghton Mifflin’s digital platform. These synergies could be worth north of $1 billion to Veritas, but Veritas is not paying for any of these in its offer to the Company’s shareholders.

There Is a Superior Alternative to Selling the Company for $21 Per Share

Rather than selling the Company for less than fair value and letting Veritas receive the value of 100% of the synergies, we believe there is a far better alternative for Houghton Mifflin shareholders. We believe long-term shareholders would be better off if the Company refinanced its existing debt, conservatively re-levered its balance sheet to 2x EBITDA and executed a Dutch tender offer between $21 and $22 per share. Assuming a repurchase at the mid-point of the range, the Company could repurchase around 20% of its shares outstanding at $21.50 per share and shrink its number of outstanding shares to around 105 million. This would lead to a levered free cash flow per share of around $2.85 at the mid-point of the billing range. If one were to apply a conservative 12x multiple to this number, the Company’s stock would be at $34 per share. We believe this would happen toward the beginning of 2024 based on the Company’s own billings forecast. This proposal would be a win-win for everyone. It would be a win for short-term shareholders looking for liquidity today, who would receive $21 to $22 per share, and it would be a win for long-term shareholders who believe in the future of the business and want to continue to own the Company.

If a majority of shareholders don’t tender into the Veritas offer and the Company remains public, we intend to nominate directors at the earliest possible opportunity if the Board doesn’t execute on the plan we have outlined in this letter.

In conclusion, we are disappointed in the Board for running a flawed process at the wrong time and accepting a price that is too low and undervalues the Company’s bright prospects. As a result, Engine will not tender its shares. We hope a majority of shareholders reach the same conclusion and long-term shareholders are able to continue to own the Company under the outstanding leadership of the current management team. We request a meeting with the Board at its earliest convenience to discuss the initiatives summarized in this letter.

Very truly yours,

Arnaud Ajdler
Managing Partner

Brad Favreau
Partner

About Engine Capital

Engine Capital LP is a value-oriented special situations fund that invests both actively and passively in companies undergoing change.

_________________________

¹ Company's Schedule 14D-9, filed on March 7, 2022.
² Company's Schedule 14D-9, filed on March 7, 2022.
³ Company's Earnings Call presentation, November 4, 2021.
⁴ Company's Schedule 14D-9, filed on March 7, 2022.
⁵ Company's Schedule 14D-9, filed on March 7, 2022.
⁶ Company's Schedule 14D-9, filed on March 7, 2022.

View source version on businesswire.com: https://www.businesswire.com/news/home/20220309005441/en/

Engine Capital LP

Arnaud Ajdler, 212-321-0048

aajdler@enginecap.com

Source: Engine Capital LP