Earlier this year, Michael Klausner and Michale Ohlrogge (K-O) posted a refutation of our paper, “No, SPACs Do Not Dilute Investors.” Their response provides no original analysis to refute our findings, merely reiterating the arguments they make in their earlier published papers. In this post, we establish why their response is hollow.

The heart of K-O’s mistake is their misallocation of all the costs of the merger of a public SPAC with a private target, referred to as a de-SPAC transaction, to the non-redeeming SPAC investors. The fact of the matter is that our paper and public proxy filings demonstrate that in reality costs are shared ratably between the SPAC and the target.

Their story goes like this: In a de-SPAC merger, all shares are valued at $10/share. This is true. But K-O then claim that the SPAC’s net cash per share (“NCPS”) – which applies all of the costs of the SPAC/de-SPAC process to the SPAC alone – reduces the cash per share from $10 to $5.70 for the median SPAC, and that is the value per share a non-redeeming shareholder can expect to receive in the de-SPAC merger. The mechanism to accomplish this, their story goes, is to inflate the value of the target above its actual value so that target shareholders get more shares in the merged company than they would otherwise.  K-O contend that inflating the value of the target can shift costs.

We start with the base example, which prevails in reality, of how pro rata sharing of costs works.  SPAC shareholders contribute their 10 shares each valued at $10/share, a total contribution of $100. The target company is worth $1,000. Target shares are also valued at $10 per share, so the target shareholders get 100 shares. Thus, before allocating costs, the combined entity is worth $1,100 with SPAC shareholders owning 10 shares (9%) and target shareholders owning 100 shares (91%). However, suppose the various costs of the merger, e.g. deferred underwriting fees and financial advisory fees, add up to $20 total, which reduces the value of the combined entity to $1,080. Each share of the combined entity is now worth $9.818 per share (not $10 anymore due to costs). The SPAC shareholder’s ownership is worth $98.18 (10 shares at $9.818 per share), and the target ownership is worth $981.82 (100 shares at $9.818 per share). The $20 of costs has been split pro rata between the SPAC shareholders, who pay $1.82 of the costs, and the target shareholders, who pay $18.18 of the costs.

K-O argue that the SPAC promoters and the target want to impose all the costs on the naïve SPAC shareholders. To accomplish this, the value of the target is inflated to $1,250 (over the real value of $1,000), so the target shareholders will receive 125 shares instead of the 100 shares previously. With the inflated value, the combined entity is worth $1,350 before costs. The SPAC shareholders still own 10 shares, but those now only represent 7.4% of the combined entity, down from 9% in the pro-rata example. The target shareholders receive 125 shares representing 92.6% of the company. Now when the $20 of costs are subtracted, the combined entity is allegedly worth $1,330 ($1,350 – $20). Since there are 135 total shares (10 for SPAC and 125 for target), the per share value of the entity is $9.852 ($1,330 / 135).

However, when public trading begins, the market recognizes the inflated value of the target and drops the value down $250 (since it was inflated from $1,000 to $1,250). So, the value of the combined entity is not $1,330 (net of costs), but rather $1,080 ($1,330 – $250), as it was in the base example and that is the correct value. But in this case, there are more shares, so the per share value is down to $8 per share ($1,080 / 135). The SPAC shareholders ownership stake is only worth $80 (10 shares at $8 per share), while the target stake is worth $1,000 (125 shares at $8 per share). Since the target company was only worth $1,000 in the first place, the target shareholders have their full value, while the SPAC shareholders absorb a $20 loss, which is exactly the amount of the costs. In this case, SPAC shareholders have now paid the full $20 of costs, while the target shareholders pay none of it.

The reality is that the target contributes shares worth $10, the non-redeeming SPAC shareholders contribute shares worth $10, and then the merged entity absorbs the costs associated with the SPAC/de-SPAC process. These costs lead in a typical SPAC to a post-merger share valuation of $9.45 as our research shows.

Past research also shows that initial SPAC shareholders are predominantly rational sophisticated institutional investors. Additionally, PIPE investors, who invest in the SPAC as part of the merger, are also sophisticated institutional investors and typically pay $9.45/share.[1] If the merger would allocate all merger costs to the SPAC by overvaluing the target, all SPAC shareholders would redeem and PIPE investors would not be available. Furthermore, many SPACs voluntarily obtain independent fairness opinions indicating that the de-SPAC merger price is based on a fair valuation of the target. These opinions serve as an additional check on any effort to inflate the value of the target.

K-O refute our view by claiming that we “assume” that “SPAC[] costs are borne pro rata by the target and the SPAC shareholders” and therefore “simply assume [our] conclusion that SPAC shareholders suffer little dilution.”  However, pro rata cost sharing is not an assumption. It is a verifiable reality. This is because all shareholders (non-redeeming SPAC investors, target, sponsor and PIPE investors) own shares in the merged company and it is the cash available to the combined company that covers the costs of the de-SPAC merger, as disclosed in the merger documents.  The merger costs are paid upon the consummation of the merger and there is never a pre-merger interim step when non-redeeming SPAC shareholders incur all of the costs of the merger and have shares that are worth K-O’s NCPS metric.

The equity value of the shares in the post-merger company are “diluted” by the cost of going public thereby reducing the equity value to $9.45/share in a typical SPAC. This reduction in equity value reflects less cash in the merged company, because this cash is paid out to cover the costs of going public. However, our research also shows that the ownership stake of the SPAC investor in the merged company actually increases by 5% due to redemptions by other investors.

So, what evidence do K-O have that target companies are systematically inflating their value to shift costs? According to K-O, the evidence is the existence of a correlation between NCPS and post-merger stock price performance. In other words, the inflated value of the target means worse post-merger stock price performance.

In their response to us, they claim that several of their regressions, using different models, find strong correlations between NCPS and post-merger share prices. Further, they wrongly accuse us of “cherry-picking one regression,” in which the statistical significance of their correlations disappears. We strongly disagree with the “cherry picking” characterization.  As we argue in our paper, there are convincing reasons why that one regression is the most relevant of all their analyses. The facts of the K-O study are key.

Fact #1 is that K-O note three confounding factors (redemptions, sponsor quality and PIPEs) that call into question the empirical relationship between NCPS and stock price that they claim to have found, even admitting that these factors “would affect the interpretation of the results….”

Fact # 2 is that K-O argue in support of using a dataset of stock prices through December 1, 2022, and use such data in all of the regressions that they report in the body of their paper, including those that they claim as their “main” results.

Given facts #1 and #2, the model that (i) controls for all three confounding factors and (ii) uses the full December 2022 dataset, which is used everywhere else in the paper, seems to be the most comprehensive and relevant of all their models. And under this model, the statistical significance of their results disappear – a fact buried in footnote 48 which discusses their results that control for redemptions, PIPEs and sponsor quality all together, conceding “[b]y the time we finalized this Essay for publication, the data, as of December 1, 2022, had shifted the results of these analyses and the cash per share figures were no longer statistically significant.”[2] [emphasis added] This critical admission is ignored in the body of the paper. Far from “cherry-picking,” we are merely highlighting what is arguably the most relevant finding from their analyses.

Of course, the K-O footnote, and subsequent response to our paper, also claims that if they truncate their dataset to end in August rather than December 2022, then the statistical significance of their results re-emerges. Footnote 48 notes that the regression coefficients are statistically significant using data as of August 2022. But that begs the question of why the August 2022 sample period is more relevant than the longer December 2022 sample period, which again is used in all their other empirical analyses. There is no mention of an August 2022 cutoff anywhere in the paper other than to dismiss the unfavorable results of the one model that controls for all relevant confounding factors. Thus, it would appear that K-O’s allegations of us “cherry picking” is a classic instance of the pot calling the kettle black.

K-O conclude by suggesting that the historically poor market performance of de-SPACs supports their claim that SPAC investors are diluted: “Only 14% of [] SPACs [that merged between 2019 and 2021] are now trading above $10.” They question, “[i]f Gulliver and Scott believe that the low pre-merger value of SPAC shares was of no relevance to post-merger share value, how do they explain these post-merger share prices? Bad luck?”

It is true that de-SPACs have moderately underperformed similar IPOs, but it does not follow that this is due to overvaluation of the target to shift costs onto SPAC investors.  It is more plausible that other features of SPAC transactions allow targets to receive high valuations, such as the historical ability of targets to disclose forward looking projections and the fact that targets tend to be younger companies with lower revenues and profits than companies that IPO. These verifiable differences between SPACs and IPOs are more plausible reasons for poor performance than the contrived “net cash per share” measure.