Over the past few years, the U.S. technology sector has been buoyed by strong market conditions, a healthy economy, and relatively easy access to financing, all of which have generally boosted its business prospects. But this benign business climate has also caused many tech companies, and notably some software firms, to increase their tolerance for risk as they embarked on a slew of debt-financed acquisitions, mergers, or leveraged buyouts. S&P Global Ratings believes this exuberance has contributed to an erosion in credit quality that could spell trouble for both debt and equity investors when the economic cycle turns.

Our analysis of the business and financial performances of our rated U.S. technology companies between 2016 and 2018 highlights the weakening credit quality given the debt-fueled growth for tech companies, most notably those involved in financial sponsor-led transactions. The substantial increase in 'B-' rated companies over that time, along with the continuing momentum for leveraged buyouts and merger and acquisition (M&A) activity, lead us to conclude that there will be a greater number of credit rating downgrades to 'CCC+' or below over the next few years.

We see that growing debt burdens can limit companies' financial flexibility and leave little room for error as they seek to integrate new acquisitions or are involved in leveraged buyouts. Moreover, in management's or their financial sponsors' eagerness to achieve a smooth business combination or significant cost savings, we also see an increased risk that some companies may underperform their original business or financial goals. Furthermore, if management teams fail to achieve their operational plans or cost savings targets, their companies could face near-term debt service issues and, if interest rates rise and credit markets deteriorate, a greater likelihood that their capital structures will become unsustainable.

Ratings On New Deals Migrating From Weak To Weaker

From 2016 to 2018, more companies accessed the debt capital market via leveraged buyout transactions, which were the primary driver of the increased number of 'B-' rated companies (see chart 1). In fact, we now see more 'B-' than 'B' ratings on new deals by financial sponsor-owned companies, despite the fact that many recent deals have involved software companies, which often have a distinct business advantage in that they can have a high base of recurring revenues.

We have witnessed record capital raised by technology-focused private equity firms, which created competition with strategic buyers for acquisitions. This, in part, has resulted in higher valuations for M&A and LBO transactions, which means more debt was issued by U.S. technology firms, many of which are also relatively small, have limited operational track records, and no experience managing their business with a heavy debt load. The higher purchase multiples means that high debt leverage is generally needed to finance transactions despite meaningful equity contributions from financial sponsors.

When calculating credit metrics for marketing purposes, companies and bankers often cite management's adjusted EBITDA, which includes various add-backs inclusive of future cost savings or deferred revenues. S&P Global Ratings analyzes companies based on its view of EBITDA that is generally different from management's, which can often overstate profitability and understate credit risk. We typically do not include these add-backs to EBITDA because we believe there are risks to achieving such add-backs. We model cost synergies in future years based on a level we view as achievable. In addition, we typically view some add-backs (such as restructuring) as a cash operating item because most companies need to restructure their operations to remain competitive and viable. While most companies achieve their targeted cost cuts, they could come at a cost of business disruptions or the lack of necessary investments to achieve their new business model. That means their forecasted margin improvements did not materialize. As we discuss below, this dynamic is responsible for many of our downgrades.

The LBO Factor

Many leveraged buyout (LBO) deals drove the increase in lower rated credits in recent years. The number of 'B-' rated LBO deals has increased from about one-third of the total U.S. tech LBO transactions in 2016 to about two-thirds in 2018 (see table 1). This has resulted from higher leverage and lower-quality credits.

Table 1

Even though management forecasts deleveraging in almost all new LBOs, credit metric improvements often take longer than expected due to ongoing restructuring or unforeseen operational underperformance. Some examples of companies whose continued restructuring activity pressured cash flows and led to downgrades include Diebold Nixdorf Inc. and Veritas Bermuda Ltd.

Chart 2: LBO Ratings And Outlooks At Transaction Close 

Speculative-Grade U.S. Tech Ratings

Of the approximately 160 speculative-grade companies we rate in the U.S. technology sector, approximately 70% are in the 'B' category (with about 40% of the total with 'B-' ratings). We currently rate 4% of speculative-grade U.S. tech credits 'CCC+' or lower, reflecting our view that those companies have unsustainable capital structures.

We see that leverage has clearly risen in new LBOs and M&A. Over the last three years, as we can see in table 2, average starting leverage for U.S. tech companies pursuing LBOs or M&A has increased about over a turn, to 7.1x in 2018 from 5.8x in 2016. Average LBO deals have increased by about two turns over the same period. We attribute this to an increasing risk tolerance by many market participants. Moreover, we can see (using our adjusted leverage metrics) that initial leverage for U.S. tech issuers engaged in LBOs, on average, surpassed the 10x mark last year.

Table 2

The software preference

While we find debt leverage to be high for financial sponsor owned hardware companies, where sales tend to be more transactional in nature, leverage is even higher at software companies, due to their subscription sales models, which have predictable revenue and low capital spending requirements that enable them to carry higher leverage. As discussed below, rising purchase multiples require significant debt leverage even for transactions where equity contribution is high. We can see in chart 4 how leverage is notably higher now for software companies than for hardware companies. There was only one LBO at a hardware firm in 2016 and 2018, and only three in 2017. By contrast, there were 10 at software companies in 2016, nine in 2017, and 11 in 2018.

Many of these highly leveraged LBO credits are smaller companies with short operating histories. The average management-adjusted EBITDA at LBO transaction close for deals rated in 2016 was approximately $295 million. In 2017 and 2018, the averages were lower, at $105 million and $218 million, respectively. We view many of these companies to have weak business risk profiles and limited revenue diversification. This makes them particularly vulnerable and less likely to withstand impact from unforeseen operational problems.

Difference In EBITDA Definition

Relying on EBITDA as defined by credit agreements can also mask deteriorating credit situations where there are significant risk of achieving the EBITDA add-backs. Our review of new issuer transactions between 2016 and 2018 shows that EBITDA add-backs represented about 20% of the starting pro forma leverage shown in deal marketing documents. That is in line with the cost-savings cap we typically find in credit agreements that allow add-backs of ongoing cost savings that management reasonably estimates are achievable within a specified period. While we expect companies to execute their intended cost reduction plans, more often than not, we find our views to incorporate a higher risk for business disruptions, which result in a slower deleveraging path.

Chart 6: Average EBITDA Add-Back And Reported EBITDA At Recent Tech LBOs 

The Equity Question

Financial sponsors often cite large equity commitments as an offset to high leverage, underscoring their confidence in a firm's business prospects. While sponsors claim higher equity checks are credit positive, we do not believe higher equity checks lead to better credit quality. In fact, the majority of the negative rating actions observed within our 2016 to 2018 data set occurred among LBO credits: Of the 85 transactions, 13 experienced negative rating actions, of which eight were at companies backed by private equity. Because valuations of companies have risen significantly, often at a much higher rate than a company's EBITDA growth, sponsors must inject more equity than what they would have in the past in order to make starting deal credit metrics more palatable to debt investors, and to ensure positive annual free operating cash flows after interest payments.

Table 3

As purchase multiples have risen, financial sponsors have increased their equity commitments alongside increased debt issuance in order to consummate the transactions. But we don't see these higher equity checks as a major factor in assessing a company's credit quality. Despite rising equity contributions, we witnessed an increase in the number of 'B-' rated companies. We conclude that a tech firm's business prospects and financial metrics are what matter most to its credit rating--not higher equity checks.

Table 4

Performance Overview: 2016-2018

So how have these companies and financial sponsors performed over time? We expected to find that companies with ambitious synergy targets and meaningful EBITDA adjustments would frequently underperform management's forecast significantly and S&P Global Ratings' base-case forecasts to a lesser extent when we assigned a final rating. Our forecasts are generally below that of management's. However, our review of 85 transactions over the 2016 to 2018 period shows that among the speculative-grade companies whose operational performance fell short of our expectations because of idiosyncratic credit issues, EBITDA add-backs and operational underperformance have a low correlation. While we saw that companies with a larger share of EBITDA add-backs or cost savings tend to underperform relative to our base-case expectation and management forecasts, the add-backs by themselves are not the cause of this underperformance. We do not view the underperformance as systemic. Instead, we believe that although companies generally achieved their cost targets, the new entities were vulnerable to business-specific disruption, underestimated restructuring costs, or overestimated growth prospects.

To help illustrate how companies could fare in the future, we analyzed the three-year performance of the 2016 cohort of debt issuers on how they performed over time. Our analysis shows that for these companies, we saw a higher number of negative rating actions on LBOs, whereas we found more positive rating actions on M&A deals. Within the 2016 data set, 15 of the 26 deals fell short of our base-case expectations and management's forecast for deleveraging at rating initiation. The remaining 11 companies performed in line to better than we had expected, but generally in line with management's expectations. Leverage at these outperformers was generally less than 5x, the EBITDA base at them was higher, and they tended to have longer operating histories than the underperformers.

Table 5

There have been 13 subsequent rating movements for the 26 transactions in the 2016 cohort, ten for the 2017 cohort, and one for the 2018 cohort. We believe the 2017 and 2018 cohorts will experience similar rating activities to the 2016 cohort once more time has elapsed from deal close.

Of the 13 transactions in the 2016 cohort that had subsequent rating actions, about half were negative and half positive. Most of the positive rating actions were among companies that engaged in M&A rather than LBO. These companies generally prepaid debt or realized business improvements faster than our expectations (but generally in line with management's expectations). Of the seven positive rating actions from the 2016 cohort (comparing current ratings to assigned ratings), three of these actions happened after one year of the deal.

Table 6

We have also seen that the number of LBO transactions in the U.S. tech sector remains steady.

For the 26 transactions we analyzed from the 2016 cohort, we took negative rating actions on five of 11 LBO issuers, mostly in the second year after rating assignment. We took a positive rating action on only one LBO issuer in the second year after rating assignment. We took positive rating actions on seven of the 15 M&A transactions, with about half occurring after the first year of ratings assignment. While our sample set is small, we find the results to be consistent with our hypothesis that companies in M&A deals are more likely to meet our expectations for deleveraging than companies in LBOs.

Chart 8: Rating Actions One, Two, And Three Years After Transaction 

The Underperfomers Tend To Be LBO Issuers While The Companies That Meet Or Exceed Our Expectations Tend To Be M&A Issuers

These are the companies from the 2016 cohort that have not met our expectations for reducing leverage. Six of the 11 companies mentioned below are LBO issuers, most of which experienced performance issuers.

Table 7

It is, however, not all doom and gloom for the tech companies involved in LBOs or M&A transactions. We identified 10 companies from our sample set of 85 transactions that generally performed in line with expectations or slightly better. We took positive rating actions on seven of these 10, upgrading three in their second year from ratings assignment while revising the rating outlooks on the remaining four to improve on what they had initially been. Only two LBO names from our 2016 cohort performed better than our expectation (Polycom Inc. and SolarWinds Holdings Inc.)

Low Recovery Prospects

Credit quality erosion is only part of the story. We have observed that the strong demand for first-lien debt by lenders has led to fewer second-lien or high yield debt issues by U.S. tech companies between 2016 and 2018. As such, first-lien debt tranches have less cushion to absorb enterprise value erosion and credit deterioration. As a result, we have assigned a lower number of '2' recovery ratings and a higher number of '3' recovery ratings (meaningful recovery of 50%-70%) to first-lien debt. LBO transactions generally continue to have some subordinated debt, but we still assigned '3' recovery ratings because of the rising debt leverage used to finance these deals and our expectation for lower recoveries, as discussed below.

Table 8

Table 9

Recovery by debt structure

We have also observed that several first-lien tranches rated for all transactions in 2018 still received a '3' recovery rating even when their capital structures include lower priority debt (typically second lien). We believe this is the case because the recovery value at default is expected to be low, and the second-lien cushion was not large enough to absorb expected losses to warrant a greater than 70% recovery value ('2' rating) on the first-lien debt. While sponsors are putting in more equity into new deals, equity is in a first loss position and we assume all equity value has eroded when running our recovery model.

Chart 9: First-Lien Recovery Ratings 

Covenant-lite deals

Below, we see the increasing number of covenant-lite deals that have come to the market in the past three years. While these covenant-lite deals tend to have lower recovery ratings, they are not subject to financial terms that could lead to technical defaults.

Potential Restructuring

In the event that some tech companies significantly underperform and eventually default, we think there will be more debt restructurings than liquidations. This is because we expect the recovery value to be highest when the companies continue to operate as a going concern, especially in software companies where hard assets are usually limited. We value substantially all issuers in the U.S. technology space using the EV/EBITDA method to ascribe recovery values. Additionally, we think looser or 'lite' credit documentation will contribute to lower recoveries in the event of payment default because companies' performance could deteriorate significantly before a covenant breach, or realize that restructuring is necessary. While companies could seek amendments and maturity extensions to alleviate shrinking covenant headroom and looming debt maturities, lenders likely won't have the chance to step in and take corrective action prior to a company's operational performance deterioration or falling short of plan in covenant-lite deals.

More 'CCC' Ratings

Credits in the 'CCC+' rating category transitioned from 'B-' typically because of poor operating performance, which led to even higher leverage and negative cash flow. These issuers, however, often do not face the risk of an imminent default or liquidity crunch, usually because of the lack of near-term debt maturities or sufficient liquidity. Nevertheless, we view negative free operating cash flows and excessive leverage as unsustainable in the long term unless favorable market and business conditions persist. Given the increased number of 'B-' credits and stretched balance sheets we expect to see more downgrades. Credit market conditions can deteriorate quickly, potentially making refinancing difficult even for weaker 'B-' rated companies that have moderate leverage and negative FOCF.

Table 10

More Leverage, More Downgrades

Over the past 12 months we have already seen a number of negative rating actions that have led to two defaults and three downgrades to 'CCC+'. By contrast, there have been only a handful of defaults by U.S. technology companies over the past decade. However, we now see a strong possibility of further downgrades in coming months, largely reflecting the higher number of transactions with elevated debt leverage.

We recognize that covenant-lite transactions lower the risk of technical default, but we find constrained balance sheets to present a significant risk to a company's ability to compete successfully. Furthermore, given the increase in first-lien debt in a company's capital structure, we believe recovery for first-lien lenders could be lower than the 80% average first-lien recovery rates we saw over the past 10 years for all U.S. corporates that defaulted (see "A 10-Year Lookback At Actual Recoveries And Recovery Ratings," published Feb. 4, 2019, and "U.S. Recovery Study: Shrinking Debt Cushions And Rising Covenant-Lite Issuance Are Beginning To Weigh On Recoveries," published Jan. 8, 2019).

Our economists now forecast a 25%-30% probability of a recession over the next 12 months (up from 20%-25% previously), and we think that in a severe downturn tech companies rated 'B' or lower would be particularly vulnerable. While many of these tech companies have no near-term refinancing needs, we believe that their best hope for staving off trouble in another recession is to successfully execute their growth plans and cost-savings initiatives and generate enough cash to reduce their leverage. There is no assurance that capital markets will be as favorable as they are now.

Appendices:

These are recent downgrades of U.S. tech companies and their ratings.

Table 11

These are the companies analyzed for this report.

Table 12

This is the rating and outlook distribution in 2016, 2017, and 2018, for the U.S. tech companies involved in M&A activity or leveraged buyouts.

Chart 11: Ratings And Outlook At Transaction Close 

Related Research

• U.S. Recovery Study: Shrinking Debt Cushions And Rising Covenant-Lite Issuance Are Beginning To Weigh On Recoveries, Jan. 8, 2019
• When The Credit Cycle Turns: The EBITDA Add-Back Fallacy, Sept. 24, 2018
• Despite Credit Risks, More Software LBOs Are On The Horizon Aug. 29, 2017
• Recovery Rating Criteria For Speculative-Grade Corporate IssuersDec. 7, 2016