articles Ratings /ratings/en/research/articles/220721-credit-faq-how-u-s-data-centers-are-navigating-inflation-and-recession-risks-12444700 content esgSubNav
In This List
COMMENTS

Credit FAQ: How U.S. Data Centers Are Navigating Inflation And Recession Risks

COMMENTS

Retail Brief: European Retailers Set Out Their Stalls For The Golden Quarter

COMMENTS

Instant Insights: Key Takeaways From Our Research

COMMENTS

Digital Assets Brief: Crypto's Trump Card

COMMENTS

Sustainability Insights: Rising Curtailment In China: Power Producers Will Push Past The Pain


Credit FAQ: How U.S. Data Centers Are Navigating Inflation And Recession Risks

Financial markets have grown increasingly volatile this year spurred by several recent developments including the rising risks of recession and inflation, as well as reverberations from the Russia-Ukraine conflict. S&P Global Ratings has been monitoring the impact of these factors on rated U.S. data centers such as Equinix Inc., Digital Realty Trust Inc., Switch Ltd., Cyxtera Technologies Inc., as well as several smaller providers and examining how they've been dealing with the precarious macro and geopolitical environments.

We believe that a heightened risk of recession, rising inflation, and rising interest rates are not likely to have an impact on ratings for scaled, diversified, investment-grade providers although smaller, more highly leveraged operators could face more pressure. We believe the data center industry is fairly recession resilient given that underlying demand trends are related to the ongoing digital transformation and consumption of data. However, the sector is not immune from macroeconomic trends, particularly cost inflation and rising interest rates.

We believe that inflation could have a meaningful impact on the profitability of some data centers over the next year. Furthermore, elevated raw material costs, coupled with rising cost of capital and possible material shortages could also impact the pace of expansion and medium-term revenue growth rates. Separately, operators with exposure to floating rate debt will experience reduced financial flexibility and lower cash flow, which could have negative rating implications for operators that are highly leveraged with weaker revenue growth rates. Below, we address frequently asked questions about input costs, hedging, and other related topics for the U.S. data center sector.

Frequently Asked Questions

How does S&P Global Ratings think a recession would impact the U.S. data centers?

We believe there's still strong demand supporting the sector driven by internet traffic growth, increased information technology (IT) outsourcing, and higher application complexity. Although we don't think the sector is insulated from an economic deceleration, we believe that the current favorable sector tailwinds will likely persist, especially considering that cloud and network providers make up a significant portion of the overall customer base for the large data center providers, such as Digital Realty and Equinix. It's less clear how the pace of growth for enterprise customers will be challenged by a recession. On the one hand, enterprise IT outsourcing to third-party data centers could decelerate if corporate decision-making regarding IT migration slows. Conversely, we believe it's possible that demand could accelerate as hybrid IT and multi-cloud solutions offer more efficiency for enterprise customers, which could come into sharper focus during a recession.

How is input cost inflation impacting the sector?

We are anticipating industrywide pressure on EBITDA margins due to inflation to range from 50 bps-300 bps in 2022 depending on the level of exposure and risk mitigants, that vary by operator. S&P Global economists forecast CPI to moderate to 4.8% growth in 2023 and 2.5% in 2024, still above the Fed's target of 2% but well below the 9% we project for the third quarter of 2022. We believe Cyxtera and Cequel Data Centers should experience more margin pressure than Equinix, Digital Realty, and Switch because of their greater proportions of unhedged power costs.

Rated providers have a wide variety of costs. For retail colocation providers, labor is often the largest expense, consisting of engineers, salespeople, and management. For example, labor consists of 37% of Equinix' overall cost structure. Like most corporations, data center companies may be forced to increase compensation to attract and retain top sales talent, which could be a modest drag on profitability over several years. For wholesale providers such as Digital Realty that tend to sell to fewer customers and don't offer managed services, selling, general, and administrative (SG&A) costs as a percentage of sales is much lower, at around 9%, limiting the risk to wage inflation.

Power is another substantial expense for all U.S. data centers but the level of exposure to rising energy prices varies by operator. Wholesale triple-net-lease contracts typically include "metered power" whereby power costs are a direct pass-through to the customer, insulating these providers from rising energy costs. For example, 84% of Digital Realty's leases floor area is separately metered for electricity consumption and gets reimbursed for over 90% of its utility costs.

However, retail colocation providers are more directly exposed to rising energy prices given that they consume vast amounts of power to run and cool servers in their facilities. Power constitutes about 27% of operating expenses for Equinix, but for others that offer more cloud and managed services, this percentage can be lower. Most retail colocation contracts don't allow for "metered power" the way wholesale triple-net-lease contracts do. However, many retail colocation contracts do have pass-through options embedded in them if power costs rise above a certain baseline threshold.

So far, the U.S. data centers have been able to pass on at least some of their energy costs to the customer either through pass-through clauses or price escalators (Equinix's range from 2% to 5%) on a periodic basis. We believe the success of the pass-through ability is related to the competitive position of each U.S. data centers. The U.S. data centers that compete more on price (e.g., Cyxtera and some smaller providers) could experience more margin compression than those with more robust platforms (e.g., Equinix, Digital Realty, and Switch).

Customer contract pass-throughs could also be limited by the following:

  • Some customer contracts contain caps on amounts to be passed through;
  • Some customers pay fixed amounts for a committed level of usage so only cost increases for overages can be passed through; and/or
  • Under certain circumstances, the U.S. data centers may decide not to pass through price increases if they believe that doing so will hurt their customer relationships.

Finally, many rated retail data center providers rent their space, potentially exposing them to rising expenses if contracts are up for renewal. For example, rising rent coupled with low utilization has hurt certain providers, such as Dawn and Cyxtera, which have been forced to close unprofitable facilities. Among players who lease most of their facilities, we would favor the U.S. data centers with longer-dated average lease terms. In contrast, U.S. data centers that own their facilities are not exposed to this risk and could benefit from higher rental income.

Switch (99% ownership), Digital Realty (85% ownership), and Equinix (60% ownership) are a few of the U.S. data centers that we would favor in the current environment because we view asset ownership as a built-in hedge against inflation.

Most of the U.S. data centers have some form of a hedging program in place to deal with rising input costs, but specific hedging strategies vary by operator. Hedging ranges from short, seasonal contracts to longer-term PPAs. This is a significant consideration, which we expect will continue to put moderate pressure on margins (100 basis points [bps]-300 bps), particularly as existing hedges roll off and new hedges become more expensive.

How are the U.S. data centers utilizing renewable energy in the current market environment?

Many data center providers have transitioned to renewable energy sources (i.e., solar, wind, biodiesel, and fuel cells), which can provide a cost advantage in an environment of elevated traditional energy prices if green energy is governed by long-term power purchase agreements (PPA) or from supplier-based green power. PPA contracts are usually fixed and negotiated with some price escalation, for as much as 30-year terms. We view facilities that purchase directly from a green source or from PPA contracts as being better positioned to deal with inflation than facilities that run on more traditional energy sources (e.g., natural gas or coal) because of the fixed and long-term nature of PPA contracts. Another source of green energy is through the purchase of renewable energy credits (REC), which doesn't provide the same amount of protection against rising energy costs. When using a REC, a data center provider still purchases electricity from the utility company, which may use a variety of energy sources in its grid ranging from wind, solar, natural gas, nuclear to traditional. By purchasing a REC, the renewable aspects are transferred to the data center provider, signaling that renewable energy was generated on their behalf, but the price of electricity can fluctuate with the prices of traditional sources.

The three leading green energy users are Switch (100%), Equinix (95%), and Digital Realty (64%) although the renewable energy type varies:

  • Equinix uses RECs for 45%; supplier green power for 37%; and virtual PPAs for 13%.
  • Digital Realty uses 40% from retail supply contracts, 33% from customer sourced renewables, and 14% from PPAs with only 2% from REC purchases.
  • Switch paid an impact fee of $27 million to NV Energy Inc. become an unbundled purchaser of energy in Nevada, allowing it to pursue open market green alternatives, which has resulted in $47 million of power-related cost savings.

This purchase of green energy can serve as a competitive advantage by helping customers meet their environmental, social, and governance (ESG) goals that could lead to lower churn rates compared to facilities that don't offer such benefits. Other providers with smaller scale in each market can make it challenging to purchase green energy regionally until the cost of RECs declines.

Is the Russia-Ukraine conflict hurting the sector? Are there supply chain issues?

So far, there don't appear to be any significant impacts from the Russia-Ukraine conflict on the supply chain dynamics for the sector although U.S. data centers have cited minor delays and extended lead times regarding their ability to source certain materials. Nevertheless, ongoing supply chain constraints could slow the pace of expansion. Given that many data center operators are heavily increasing capital expenditures (capex) to meet demand, we believe the credit impact would be minimal as free operating cash flow (FOCF) could temporarily expand with less capital spending.

Key commodities for new data center construction include steel, concrete, copper, aluminum, and lumber, all of which have seen prices surges over the past year and have only recently abated from local highs (see below price chart). We believe U.S. data centers that have strong relationships with their suppliers can negotiate longer duration of contract terms for such materials. Larger U.S. data centers may be able to negotiate better pricing because they're buying at scale, which would partially offset price increases. Digital Realty and Equinix also have high credit quality, which may help gain preferential treatment with suppliers. A similar dynamic occurs for finished parts including power distribution and cooling and Heating, ventilation, and air conditioning (HVAC) systems. These pieces of equipment are typically purchased from original equipment manufacturers (OEMs) and pricing is relationship-dependent.

Commodities Most Used By Data Centers

image

We're seeing more U.S. data centers, like Cyxtera, continuing to develop their smart cabinet and bare metal technology. These products allow for enterprises to seamlessly toggle space and power resources as more or less is needed and have the potential to disrupt the traditional colocation sector in the long run. Should supply chain constraints become an issue, we would expect these products could see greater demand among prospective customers.

While there doesn't appear to be any supply constraints on commodities because of the conflict, the U.S. data centers have cited modest price increases on capital purchases in recent months. However, these purchases typically coincide with new sales activity, and the U.S. data centers can adjust pricing accordingly to capture increased capital costs. Digital Realty's major equipment procurement contracts typically have a 2-3 year initial term, which somewhat insulates them from short-term price fluctuations.

On the demand side, to the extent that customers experience delays in receiving their equipment, there could be some damage, which could slow the timing of certain deployments and diminish utilization ability. While cyber-attacks have become more of threat due to the Russia-Ukraine conflict, in our view, the sector isn't seeing any major upticks in demand for their purpose-built facilities. Nonetheless, certain providers have noted that customer conversations around the topic of cybersecurity have increased, and we would expect this trend to be a net positive for U.S. data centers for at least the next few years.

How will rising rates impact the U.S. data centers?

For most U.S. data centers at the lower end of the ratings spectrum, their capital structures contain senior secured floating rate debt with no fixed rate debt. This could erode FOCF for these companies, which tend to be smaller, more highly leveraged and operate with less financial flexibility. We believe Cyxtera is most exposed to rising rates given their exposure to floating rate debt and absence of an interest rate hedging program. While Cequel and Dawn both have a significant amount of floating rate debt, they employ robust hedging programs that should provide near-term protection.

As the period of low interest rates comes to an end, the decision to finance expansion projects with a higher-cost capital is a trickier one. We do not expect a material cutback in spending due to higher cost of capital given there's still strong underlying demand for data center services and cloud-based IT infrastructure, but certain projects could be delayed or cancelled, particularly if they're built speculatively. With choppy equity markets potentially reluctant to fund growth, we believe this could mean the U.S. data centers see leverage tick up in the coming quarters if expansion is funded with debt. If the cost of capital becomes too high, such that capex slows, financial profiles of lower rated entities could actually improve if FOCF grows and it applied toward debt reduction.

How sensitive are rated operators to rising costs and interest rates?

We believe that most providers--particularly the scaled data center providers--have ample cushion in ratings to accommodate even the most pessimistic scenarios.

Below is a scenario analysis to show how each of the rated U.S. data centers credit metrics and ratings would be influenced by varying degrees of inflation and interest rates in 2022.

Data Center Subsector Stress Test
Company Rating as of July 21, 2022 Outlook Downgrade or outlook revision threshold

Equinix Inc.

BBB Stable >5x leverage

Digital Realty Trust Inc.

BBB Stable >7.5x leverage

Cyxtera Technologies Inc.

B- Stable Capital structure is potentially unsustainable

Cequel Data Centers L.P.

B- Stable Capital structure is potentially unsustainable

Dawn Acquisitions LLC

CCC Negative Liquidity continues to deteriorate such that a default or restructuring appears to be inevitable within six months
Inflation Scenario COGS/Revenue EBITDA Impact Leverage FOCF Breach Downgrade Trigger?
Equinix Low (0 bps) 54% $0 million 4.2x $186 million No
Medium (100 bps) 55% $72.8 million 4.3x $128 million No
High (250 bps) 57% $182 million 4.5x $40 million No
Very High (500 bps) 59% $364 million 4.8x Negative $106 million No
Digital Realty Low (0 bps) 40% $0 million 6.5x $434 million* No
Medium (100 bps) 41% $45 million 6.6x $389 million* No
High (250 bps) 44% $158 million 7.0x $275 million* No
Very High (500 bps) 45% $225 million 7.2x $207 million* No
Cyxtera Low (0 bps) 52% $0 million 7.5x Negative $14 million No
Medium (100 bps) 53% $8 million 7.7x Negative $22 million No
High (250 bps) 55% $18 million 8.1x Negative $34 million No
Very High (500 bps) 57% $36 million 8.8x Negative $53 million Possibly
Cequel Low (0 bps) 52% $0 million 6.7x $23 million No
Medium (100 bps) 53% $5 million 6.9x $20 million No
High (250 bps) 54% $14 million 7.2x $11 million No
Very High (500 bps) 57% $24 million 7.7x Negative $2 million Possibly
Dawn Low (0 bps) 75% $0 million 10.3x Negative $54 million No
Medium (100 bps) 76% $2 million 10.6x Negative $56 million Ongoing evaluation based on equity infusions
High (250 bps) 78% $4 million 11.1x Negative $59 million Ongoing evaluation based on equity infusions
Very High (500 bps) 80% $9 million 11.9x Negative $65 million Ongoing evaluation based on equity infusions
Interest Rates Scenario Base Rate Interest Expense EBITDA / Interest Coverage FOCF / Debt Breach Downgrade Trigger?
Equinix Lower (50 bps below) 2.50% $507.3 million 5.9x 1.8% No
Current Base (0 bps above) 3.00% $510.7 million 5.8x 1.8% No
Medium (25 bps above) 3.25% $512.8 million 5.8x 1.7% No
High (100 bps above) 4.00% $517.4 million 5.8x 1.7% No
Digital Realty Lower (50 bps below) 2.50% $337.4 million 5.2x 11.0% No
Current Base (0 bps above) 3.00% $339.5 million 5.2x 11.0% No
Medium (25 bps above) 3.25% $340.6 million 5.2x 11.0% No
High (100 bps above) 4.00% $343.9 million 5.2x 10.9% No
Cyxtera Lower (50 bps below) 2.50% $149.7 million 5.2x 0.2% No
Current Base (0 bps above) 3.00% $154.5 million 1.6x 0.0% No
Medium (25 bps above) 3.25% $156.9 million 1.6x (0.1%) No
High (100 bps above) 4.00% $164.1 million 1.5x (0.5%) Possibly
Cequel§ Lower (50 bps below) 1.50% $77.6 million 2.1x 2.9% No
Current base (0 bps above) 1.50% $77.6 million 2.1x 2.9% No
Medium (25 bps above) 1.50% $77.6 million 2.1x 2.9% No
High (100 bps above) 1.50% $77.6 million 2.1x 2.9% No
Dawn† Lower (50 bps below) 2.18% $30.5 million 1.9x 0.0% No
Current Base (0 bps above) 2.18% $30.5 million 1.9x 0.0% Ongoing evaluation based on equity infusions
Medium (25 bps above) 2.18% $30.5 million 1.9x 0.0% Ongoing evaluation based on equity infusions
High (100 bps above) 2.18% $30.5 million 1.9x 0.0% Ongoing evaluation based on equity infusions
Leverage defined as S&P Global Ratings-adjusted debt to EBITDA. bps--Basis points. N/A--Not applicable. FOCF defined as S&P Global Ratings-adjusted free operating cash flow. Inflation scenarios were generated keeping 2022 interest rates constant at 2.50% base rate. Interest rate scenarios were generated keeping inflation constant at 100 bps of margin impact. *Digital Realty's FOCF is calculated after development spend net of dispositions. §Cequel is fully hedged on its term loan at the above fixed rate under all scenarios. †Dawn is fully hedged on its term loan at the above fixed rate under all scenarios.

This report does not constitute a rating action.

Primary Credit Analysts:Shaun Epstein, New York 1 (212) 438 0232;
shaun.epstein@spglobal.com
Chris Mooney, CFA, New York + 1 (212) 438 4240;
chris.mooney@spglobal.com
Secondary Contact:Fernanda Hernandez, New York + 1 (212) 438 1347;
fernanda.hernandez@spglobal.com

No content (including ratings, credit-related analyses and data, valuations, model, software, or other application or output therefrom) or any part thereof (Content) may be modified, reverse engineered, reproduced, or distributed in any form by any means, or stored in a database or retrieval system, without the prior written permission of Standard & Poor’s Financial Services LLC or its affiliates (collectively, S&P). The Content shall not be used for any unlawful or unauthorized purposes. S&P and any third-party providers, as well as their directors, officers, shareholders, employees, or agents (collectively S&P Parties) do not guarantee the accuracy, completeness, timeliness, or availability of the Content. S&P Parties are not responsible for any errors or omissions (negligent or otherwise), regardless of the cause, for the results obtained from the use of the Content, or for the security or maintenance of any data input by the user. The Content is provided on an “as is” basis. S&P PARTIES DISCLAIM ANY AND ALL EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, ANY WARRANTIES OF MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE OR USE, FREEDOM FROM BUGS, SOFTWARE ERRORS OR DEFECTS, THAT THE CONTENT’S FUNCTIONING WILL BE UNINTERRUPTED, OR THAT THE CONTENT WILL OPERATE WITH ANY SOFTWARE OR HARDWARE CONFIGURATION. In no event shall S&P Parties be liable to any party for any direct, indirect, incidental, exemplary, compensatory, punitive, special or consequential damages, costs, expenses, legal fees, or losses (including, without limitation, lost income or lost profits and opportunity costs or losses caused by negligence) in connection with any use of the Content even if advised of the possibility of such damages.

Credit-related and other analyses, including ratings, and statements in the Content are statements of opinion as of the date they are expressed and not statements of fact. S&P’s opinions, analyses, and rating acknowledgment decisions (described below) are not recommendations to purchase, hold, or sell any securities or to make any investment decisions, and do not address the suitability of any security. S&P assumes no obligation to update the Content following publication in any form or format. The Content should not be relied on and is not a substitute for the skill, judgment, and experience of the user, its management, employees, advisors, and/or clients when making investment and other business decisions. S&P does not act as a fiduciary or an investment advisor except where registered as such. While S&P has obtained information from sources it believes to be reliable, S&P does not perform an audit and undertakes no duty of due diligence or independent verification of any information it receives. Rating-related publications may be published for a variety of reasons that are not necessarily dependent on action by rating committees, including, but not limited to, the publication of a periodic update on a credit rating and related analyses.

To the extent that regulatory authorities allow a rating agency to acknowledge in one jurisdiction a rating issued in another jurisdiction for certain regulatory purposes, S&P reserves the right to assign, withdraw, or suspend such acknowledgement at any time and in its sole discretion. S&P Parties disclaim any duty whatsoever arising out of the assignment, withdrawal, or suspension of an acknowledgment as well as any liability for any damage alleged to have been suffered on account thereof.

S&P keeps certain activities of its business units separate from each other in order to preserve the independence and objectivity of their respective activities. As a result, certain business units of S&P may have information that is not available to other S&P business units. S&P has established policies and procedures to maintain the confidentiality of certain nonpublic information received in connection with each analytical process.

S&P may receive compensation for its ratings and certain analyses, normally from issuers or underwriters of securities or from obligors. S&P reserves the right to disseminate its opinions and analyses. S&P's public ratings and analyses are made available on its Web sites, www.spglobal.com/ratings (free of charge), and www.ratingsdirect.com (subscription), and may be distributed through other means, including via S&P publications and third-party redistributors. Additional information about our ratings fees is available at www.spglobal.com/usratingsfees.

 

Create a free account to unlock the article.

Gain access to exclusive research, events and more.

Already have an account?    Sign in