This is our third annual analysis of EBITDA addbacks. It demonstrates a continuing trend of most issuers being unable to even come close to achieving the earnings, debt, and leverage projections presented in their marketing materials at deal inception. Of particular interest was whether EBITDA adjustments did result in a more accurate picture of future earnings. The answer appears to be no.

Building on the earlier data, the latest data cohort covers a sample of M&A and LBO transactions originated in 2017. Our analysis consists of two main components:

• First, we look at the validity and accuracy of the addbacks that companies built into their forecasts. Like our previous analyses, we compared issuers' projected adjusted EBITDA at deal inception with actual reported EBITDA for the two calendar years following the year of origination. Given the difficulty and limited visibility in the earnings breakout, we did not parse out the specific components of addbacks to determine individual line-item realizations. A portion of the delta between management projected and actual reported EBITDA could be attributable to factors such as unmaterialized growth or unanticipated operating issues. Nevertheless, those factors reflect the overly optimistic assumptions often used to build marketing financial projections.
• Second, we examined a broader data set of deals originated from 2015-2019 to measure the magnitude and distribution of company projected addbacks across major categories over time. This allows us to track and quantify the evolution of addbacks.

While the findings herein are an excellent reminder of the perils of optimistic forecasts, our ratings are based on our projections of a company's expected earnings, capacity and appetite for debt repayment, and our view of issues like expected synergies or cost efficiencies.

Over the past five years, most U.S. issuers that were assigned a speculative-grade rating were initially rated in the 'B' category. If we took the marketing leverage presented to us and rated to pro forma addbacks, projected earnings, and debt reduction, our initial issuer credit ratings would likely be higher, and then we would subsequently lower them as actual results are reported. We did not observe such a downward rating transition pattern across our sample. All told, marketing leverage and the language around addbacks--as defined in debt agreements--are not determinants of our view of credit risk (other than in assessing covenant headroom when reviewing debt instruments containing financial maintenance covenants).

Part 1: The Validity And Accuracy Of EBITDA Addbacks

Do addbacks present a more realistic picture of future profitability and risk, and do companies typically hit their forecast?

As noted in our previous studies on addbacks, management teams and deal arrangers continue to be creative in presenting what qualifies as an addback. This results in an increase in both the number, types, and--ultimately--magnitude of adjustments. In some of these cases, S&P Global Ratings views expanding the definition of management-adjusted EBITDA as an artificial deflation of leverage or an artificial inflation of profitability that contributes to understated valuation multiples. The absence of a uniform and commonly accepted definition of EBITDA is the key issue here. In practice, it is and has always been a negotiated definition, varying from agreement to agreement.

Summary of findings

In this study, we found once again that both anticipated EBITDA and deleveraging efforts fell materially short of issuer projections for the two years that we tracked companies' performance after transaction origination (see Table 1). We repeated the performance gap analysis for M&A and LBO transactions originated in 2015, 2016, and 2017. Specifically, our analysis of the 2017 cohort showed that on a median basis the magnitude of the misses improved modestly but the average miss remained substantial at 2.6 turns higher than management forecasts for 2018 (the first full year of performance since syndication), growing to 2.7 turns in 2019.

Over the two-year span, more than half of the companies missed their EBITDA targets by at least 25% (still lower than 65% for the 2016 cohort), and about a quarter of them failed to meet their projections by more than 50% by the end of the second year. The median miss on earnings over Years 1 and 2 were 32% and 30%, respectively, which is a marginal improvement over the 2015 and 2016 vintage cohorts. For the 2015 cohort, the median misses for Years 1 and 2 were 33% and 39%, respectively. For 2016, the misses were 30% for Year 1 and 35% for Year 2. This perhaps suggests that the legitimacy of addbacks is likely questioned by investors (see Table 2). We chose median metrics for comparison because we observed a fair amount of variation within in each cohort and across the three sets of cohorts.

Table 1

Table 2

Table 3

About the sample:  To assess the validity of addbacks, we compared marketing EBITDA presented at deal inception with the actual reported EBTIDA. We did the comparison on the aggregate level, given the difficulty in evaluating the various individual components of addbacks. For example, a company often does not disclose the actual achievement of a particular type of cost savings in its financials. We include two years of actual performance data--allowing time to gauge whether the company was able to achieve anticipated synergies--to permit certain addbacks to roll off.

Further, just like our review the last two years, we eliminated companies that underwent a material M&A or LBO transaction after the initial transaction (for example, a subsequent sponsor-to-sponsor LBO in Year T+2). This enabled us to remove distortion arising from subsequent transformative events (new debt issuance, earnings colored by subsequent acquisitions, etc.), rendering initial projections irrelevant. It also let us cleanly track the reported EBITDA, debt, and leverage versus what was projected by these companies. Lastly, because management projections are confidential, we cannot disclose any company names.

EBITDA still fell well short of management projections...

As opposed to seeing a convergence between management's projected and reported EBITDA, which would be expected as companies realize their anticipated earnings and as one-time items fall away and synergies are achieved, we saw a divergence. Moreover, this divergence grew between Years 1 and 2 in all but one case. The divergence indicates unmaterialized growth projections, operating challenges, and unrealized synergies or unattained cost savings. There is a clear bias to overestimate EBITDA in management projections.

The median EBITDA in our 2017 cohort was was roughly 30% below projected in each of the first two years (see Table 1). By the end of the second year, more than half (54%) of the companies still underachieved their original target by at least 25%. Although the magnitude of the misses improved modestly, the average earnings miss remained substantial at 27% in Year 1 and 30% in Year 2. This compares with 35% and 35%, respectively, in the prior year's (2016) cohort.

Chart 1

Table 4

...coupled with a failure to reduce debt as projected ...

Failure to meet projected debt levels also contributed to the large miss of management projected leverage. Virtually all issuers present a deleveraging story to the market at deal inception, with surplus cash swept to reduce debt in management projections. Ironically, almost all issuers contend that their projections are conservative during management presentations prior to deal launch. While the 2017 cohort of transactions showed marginal improvement in projecting earnings, it was more than offset by a significantly larger miss on projected debt levels. The median miss for the 2017 cohort was 11% in Year 1, growing to 25% in Year 2 following origination (see Table 5). This compares to 3% and 11%, respectively, for the 2016 cohort and 1% and 12% for deals done in 2015.

For the 2017 data set, companies were roughly in line with projected outstanding debt in Year 1. However, by the end of Year 2, they on average retained on their balance sheets 12% more debt than anticipated. Standing out were eight companies (all but one of which was owned by private equity sponsors) under-projected debt by over 50% in Year 2. In all cases, we assigned a rating in the 'B' category at the deal inception. This reflects our view of notable performance and execution risk over the two-year horizon on the basis of their inherently weak financial strength and vulnerability to adverse business conditions. When coupled with execution risk, this implies a great amount of performance volatility in the medium term.

In short, companies' intentions to apply surplus cash to pay down debt appears to be infrequently executed: We noticed that companies rarely, if ever, pay down debt to the extent indicated in marketing materials. All three vintages displayed a similar pattern: roughly two-thirds of companies kept debt levels in check (exceeded or within 10% of their targets for projected debt) in the first year following origination. That share quickly deteriorated to an average of 43% by the end of the second year. For comparability, we netted reported cash balances against reported debt to compute both debt and leverage divergence.

Chart 2

Table 5

…resulting in actual leverage far above initial projections.

As a result, there is a material discrepancy between projected leverage and reported leverage across the aggregate data set. On both ends, we see a company's projections become increasingly aspirational, building a significant leverage cushion and presenting a case that does not necessarily represent actual credit realities. By averaging the median gap across the three vintages, companies under-projected leverage by an average of over two turns (2.3x) in the first year, cumulating three turns (3.1x) by the end of Year 2 (see Table 6).

Chart 3

Table 6

Part 2: The Magnitude And Composition Of EBITDA AddBacks

The data set for this review includes 365 M&A and LBO transactions originated between 2015 and 2019 with deal sizes exceeding $50 million. It only includes S&P Global Ratings-rated transactions and is limited to those where management provided us with a detailed bridge from reported EBITDA to marketing EBITDA. Our final sample contains a total of 210 M&A transactions and 155 LBO transactions; of the total, 85% by deal count were rated in the 'B' category at inception, with the remaining 15% in the 'BB' rating category. With the expansion of the data set to include a new cohort, the share of 'B' category ratings has grown in tandem, reflecting the erosion of credit quality in the broader leveraged finance market. In addition, we have added a comparison of sponsored versus nonsponsored transactions; three quarters of the transactions in our sample were sponsored.

Chart 4A

Chart 4B

Chart 4C

Lower leveraged loan activity last year came with significantly higher addbacks.  We measured the magnitude of addbacks both as a percentage of management's marketing EBITDA and pro forma last-12-month (LTM) EBITDA, both as presented at transaction inception. On average over the past five years, addbacks made up over 28% of marketing EBITDA and close to 54% of LTM reported EBITDA (see Chart 5). Over the period the forward-looking measure (addbacks as a percent of marketing EBITDA) has grown marginally each year, exceeding 30% in 2018 and beyond from 24% in 2015.

Meanwhile, average addbacks as a share of LTM EBITDA declined steadily between 2015 and 2017 before jumping to 57% and 66% in 2018 and 2019, respectively. The 2015 average is inflated by its unusually high concentration of small health care and tech companies, which have displayed tendencies toward more aggressive adjustments. In 2015, over half of these companies marketed addbacks outsizing their LTM EBITDA. Excluding these deals, the average would have dropped to 29% in 2015. The jump in 2019 is somewhat expected given the surge of M&A activity in that year, which skewed the backward-looking LTM measures upward and made them out of sync with the post-acquisition business.

Across the five year sample, a large percentage of the average is weighted toward 'B' rated issuers and has been steadily increasing. We found that regardless of transaction type, 'B' category credits led their higher-rated 'BB' counterparts in the average amount of adjustment.

Chart 5

Synergies and cost savings made up about a third of total addbacks.  Expected synergies and cost savings continue to be largest component of add-backs. Chart 7 sorts the general add-back adjustments into six broad categories. On average and in each year, synergies/cost savings led other adjustment types. It peaked in 2016 at close to 39%, with a five-year average of 29%. Synergies are often the most difficult of the common addbacks to accurately project. As mentioned earlier, we rarely factor the full amount of management anticipated synergies into our projections. Rather, we have detailed discussions with management teams and their advisors around expected synergies and make our own assessment as to what we believe to be achievable. And it often depends on the source of synergy and, when relevant, whether a company has demonstrated it can realize similar synergies or cost savings from past transactions. While some are easier to execute--such as eliminating overlapping corporate overhead to achieve labor savings--others fall outside of management's control. Pro forma saving on procurement offers one example, as it takes contract negotiations with various third-party vendors. Lastly, some synergies are costly to implement, requiring an upfront expense, such as with severance pay.

Restructuring costs are another area of disparity in treatment. We generally treat restructuring charges as operating costs because we believe most companies need to restructure their operations to adapt to changing environments and remain competitive and viable. Similarly, management fees also constitute a cash operating cost and are treated as such in our analysis. For this reason, we do not add back restructuring costs or management fees in our calculation of adjusted EBITDA. In addition, there is the consistent body of data that demonstrates how far off companies' original assumptions tend to be about the future realization of addbacks.

Chart 6

Media and tech transactions had the most add-backs

The technology and media/entertainment/leisure sectors had the most addback-inflated EBITDA when comparing the five-year average of total add-backs to company marketing EBITDA at deal inception at 37% and 34%, respectively. At the other end of the spectrum, restaurants/retailing (22%), business and consumer (19%), and forest products (19%) had addbacks well below the corporate wide average of 28%.

Table 7

Chart 7

Table 8

'B' versus 'BB' rated companies:  The number of companies rated in the 'B' category continues to increase. These companies have consistently underperformed 'BB' category credits in projecting earnings. We believe this is likely because addbacks for 'BB' category credits were materially lower, so the projections relied less on achieving pro forma synergies and other future benefits. Further, the need for addbacks to make a deal appear attractive to the market is likely lower for 'BB' rated companies because their pro forma leverage is typically lower, so it is possible that the addbacks were less aggressive or aspirational. In addition, we could offer an intuitive view that lower-rated credits tend to be smaller and have higher earnings volatility, making projections more difficult. Also, financial sponsor ownership is more common among lower rated entities than among those in the 'BB' category.

On average, 'B' category credits reported leverage 2.9 turns higher than projected in 2018, with the gap widening to 3.6 turns in 2019. 'BB' category credits performed significantly better than 'B' category credits, missing by 2.5 turns in 2018 and improving to 1.5 turns in 2019. This analysis further reinforces the significant credit disparity between 'B' and 'BB' credits.

Table 9

Chart 8

Chart 9

LBO versus M&A transactions:  Consistent with our prior studies, LBO and M&A transactions are comparable in the amount of addbacks as a percentage of marketing EBITDA, at 27% and 30%, respectively. However, the distribution of addbacks differs. As one would expect, M&A transactions showed above-average addbacks for synergies and cost savings as these are often a selling point of the transaction, accounting for about 31% of addbacks versus 26% for LBOs.

Compared with prior years, the 2017 cohort of transactions shows a material gap in leverage projection performance between M&A and LBO transactions. For the 2016 cohort, there was not a pronounced difference in the quality of management projections; both proved unreliable, with the discrepancy between management projected and reported ranging between 2.7-3.6 turns of leverage across both universes. However, the gap widened materially in the 2017 cohort, with M&A deals missing leverage targets by 2.3 turns in 2018 and 2.6 turns in 2019 compared with LBO deals missing by 3.5 and 4.1 turns. For comparison, within our financial risk categories, the difference between the mid-points of two different categories (significant and aggressive, for example) is 1.0 turn of leverage.

Table 10

Chart 10

Chart 11

Sponsored vs. non-sponsored transactions:  Our five-year study on the magnitude and composition of addbacks show that sponsored transactions tend to be more aggressive with addbacks versus nonsponsored deals--but not by a significant margin. The five-year average for sponsored deals was 29.3% versus 25.6% for nonsponsored. Nonsponsored deals were generally about 25% each year with limited fluctuations. Sponsors, on the other hand, padded marketing EBITDA with more adjustments in 2018 and 2019, both in excess of 30%, likely driven by the increase in lower-rated deals given accommodative capital markets. Of the 365 transactions in our data set, 275 were sponsored, and 90 were nonsponsored.

Chart 12

Reverting to the data set in Part I, Table 11s and 12 show that sponsored transactions significantly underperformed nonsponsored in the accuracy of their projections at deal inception. For the 2017 cohort, the median leverage miss for sponsored deals in 2018 was 2.8 turns, growing to 3.2 turns in 2019. For nonsponsored deals the median miss was 1.3 turns in 2018 and 1.6 turns in 2019.

Table 11

Table 12

Conclusion

Consistent with our previous studies, this report demonstrates that EBITDA addbacks at deal inception continue to be substantial and overstated. Default rates are on the rise, and S&P Global Ratings Research expects the S&P/LSTA Leveraged Loan Index lagging-12-month default rate (by number of issuers) to increase to 8% by June 2021 from 4.6% as of September 2020. We will closely monitor the default and recovery performance of entities with substantial EBITDA addbacks to gauge the relative recovery performance of addback-laden companies versus those that are considerably less aggressive.

It will be interesting to compare initial transaction valuations versus bankruptcy emergence valuations for companies that were very aggressive versus those where addbacks were an insignificant percentage of marketing EBITDA. Our studies over three years of management projection performance show that aggressive EBITDA adjustments have understated leverage and purchase price multiples. Our study led us to several conclusions consistent with our previous studies of the 2015 and 2016 cohorts: Marketing EBITDA, including addbacks, continues to be an unreliable indicator of future EBITDA, and companies increasingly overestimate debt repayment. The combination of these factors tends to understate future leverage and credit risk, and addbacks present incremental credit risk in the form of future event risk covenants that rely on EBITDA and may provide additional flexibility under negative covenants and restricted payments (dividends, debt, and lien allowances).