In light of recent questions from collateralized loan obligation (CLO) managers and trustees on the CDO Monitor test that is applicable to cash flow CLO transactions, S&P Global Ratings is providing updated guidance on several topics regarding CDO Monitor with the aim to help CLO market participants better understand the CDO Monitor Test.

We rate cash flow CLO transactions backed by actively managed collateral portfolios using either a "stable quality" or a "stressed quality" approach. We apply the "stable quality" approach to CLOs that commit to use S&P Global Ratings' CDO Monitor (CDOM) Test as part of the reinvestment conditions to assess the impact of portfolio asset trading on a transaction's rated liabilities (tranches). Transaction documents typically cite this test as one of the collateral quality tests to be measured from the effective date and during the reinvestment phase.

The CDOM Test provides an indication of whether changes to a portfolio are generally consistent with the transaction parameters we looked at when we initially assigned ratings to the CLO notes. CLO participants (collateral managers, trustees, and/or collateral administrators) typically run CDOM on the effective date and during the reinvestment period (before and after asset purchases). CDOM is used to track the evolution of a portfolio, ultimately to determine whether a change to the portfolio due to reinvestments complies with the CDOM Test.

If the CDOM Test is failing, transaction documents typically require the manager to maintain or improve the results of the test with subsequent trades until the test is brought back into compliance.

CDO Monitor Basics

Most cash flow CLOs allow for the reinvestment of principal cash received from assets (typically those that have been sold, those that have prepaid or paid down, or recovery proceeds received from defaulted assets) during a transaction's initial specified reinvestment period. Such transactions generally have asset eligibility criteria and contractual provisions (portfolio profile tests, collateral quality tests, and reinvestment criteria) that govern the types of trading allowed and the requirements for maintaining the CLO dynamics (changes to asset portfolio and structural enhancements) within certain boundaries.

Pursuant to our rating framework, our analytical approach for rating a CLO may be based on the manager either:

• Managing the transaction to maintain the portfolio's original characteristics (the stable quality approach), or
• Managing the transaction within the eligibility criteria and contractual provisions of the governing documents (the stressed portfolio approach).

In the stable quality approach, our credit and cash flow analysis focuses primarily on the characteristics of the actual portfolio. In the stressed portfolio approach, we would assume a worst-case (yet document-compliant portfolio) in our analysis. To date, all post-crisis CLOs we have rated (CLO 2.0s) have adopted the stable quality approach in both Europe and the U.S.

To apply the stable quality approach in rating a CLO, we must be able to track the collateral manager's reinvestments, such that those reinvestment decisions are in line with the modeling assumptions we used when we assigned our initial ratings. We thus require that the manager commit, in the transaction documents, to run the CDOM Test when reinvesting proceeds during the reinvestment phase of the CLO.

In summary, CDOM is a tool that tracks changes to a CLO transaction during its reinvestment period by capturing the evolution of the credit quality of the portfolio and structural enhancements (see table 1).

Table 1

What CDOM Test Does

The CDOM Test is one of the collateral quality tests we require for monitoring purposes when rating note issues in a CLO transaction using the stable approach. It is typically applicable to the junior-most term note in the CLO with the highest initial rating (excluding class X notes and combination notes). This normally means the CDOM Test will be run on the junior-most tranche/note with an initial 'AAA' rating. If the CLO hasn't issued 'AAA'-rated notes, the CDOM Test will be applicable to the next highest rating category in the transaction's capital structure ('AA', 'A', etc.).

Assessing portfolio credit quality with SDRs

CDOM generates scenario default rates (SDRs), which represent the expected levels of defaults in a CLO portfolio under various stress scenarios based on the portfolio's characteristics and our default probability and correlation assumptions defined in our criteria (see "Global Methodology And Assumptions For CLOs And Corporate CDOs," June 21, 2019). For example, the 'AAA' stress scenario assumes an extreme level of stress, similar to that experienced during the Great Depression (widespread collapse of consumer confidence, major dislocations in the financial system, and global economic decline). As a result, the portfolio is assumed to experience a higher level of defaults in the 'AAA' stress scenario relative to stresses commensurate with lower rating levels.

The SDRs that CDOM produces at a given rating level may increase or decrease with changes in the credit quality of the underlying portfolio, including changes in the rating and maturity composition of the assets and changes in obligor, industry, and region concentrations. An increase in the SDR--caused by a deteriorated rating profile, longer maturity profile, or lower diversification--may indicate a higher potential expected default rate, or an overall riskier portfolio of assets.

Assessing tranche credit support (structural enhancement) with BDRs

The break-even default rate (BDR) represents an estimate of the maximum amount of defaults a tranche can withstand and still be able to meet its contractual payments. Notes with higher subordination/credit enhancement typically have higher BDRs.

The CDOM Test compares the SDRs with the tranche-specific BDRs to assess whether the test is passing or failing (i.e., whether the BDR is greater than/equal to or lower than the SDR). The BDR tracks changes in structural enhancements resulting from changes to the portfolio, including the amount of excess spread available and changes in recovery expectations. Greater excess spread or higher recovery expectations, all else being equal, will typically boost the BDR.

CDOM then adjusts this BDR upward or downward to account for changes in the par value of the portfolio relative to the balance at origination. An increase in par will increase the BDR, while a decrease in par (typically resulting from defaults) will lead to a decrease in the BDR.

If the SDR exceeds the corresponding BDR, the CDOM Test will be deemed to be failing for the CLO.

CDO Monitor Versions--Model And Non-Model

The CDO Monitor Test can be run using either of two approaches: a model-based approach (CDOM Model) or a formula-based approach (CDOM Non-Model).

Model version

The model-based version of the CDOM Test (like the formula version) consists of two components: the SDRs (that CDOM Model generates) and the BDRs (that the model generates, but also adjusts). We share the BDRs that are specific to each CLO traches (in the form of an input file) with the trustee and/or manager when we assign new ratings or when the CLO is close to going effective.

This input file houses up to 10,000 BDRs associated with a combination of some or all of the following parameters:

• Weighted average spread (WAS),
• Weighted average recovery rate (WARR), and
• Weighted average coupon (WAC).

This can be thought of as a matrix of up to 10,000 BDRs using a combination of WAS, WAC (if applicable), and WARR values (which are specific to each tranche rating category).

CLO Participants can download CDOM Model by registering on our website at www.sp.sfproducttools.com/sfdist/. Click on the "Analytical Models & Tools" tab, scroll down to the CDO Evaluator section, and click on the link for the applicable version of the model.

Formula version

CDOM Non-Model (a.k.a., the formula version), which we introduced in 2014, is an alternative to the model-based approach. CDOM Non-Model provides results that are analytically consistent with our rating framework, but more transparent and simpler to implement than the model-based approach. This is why CDOM Non-Model has been clearly the more popular option in both the U.S. and EMEA.

CDOM Non-Model uses a regression formula to determine the BDR result and rests on a foundation of six portfolio benchmarks (see chart below) that together determine collateral credit quality. When expressed using a simple regression formula universal to all standard cash flow CLOs we rate globally (i.e., one formula for 'AAA' CDOM Non-Model Tests and another for 'AA' CDOM Non-Model Tests), this formula produces a close approximation of the SDR that CDOM Model would generate for CLO portfolios.

The SDR is calculated by entering the portfolio into a spreadsheet (see "How To Build Your Own CDO Monitor E8 (Non-Model Version)", June 21, 2019) and then applying the portfolio-specific benchmarks to the relevant equation below (SPWARF--S&P Global Ratings' weighted average rating factor; DRD--Default rate dispersion; ODM--Obligor diversity measure; IDM--Industry diversity measure; RDM--Regional diversity measure; WAL--Weighted average life).

• CDO Monitor Non-Model SDR (AAA) = 0.247621 + (SPWARF/9162.65) - (DRD/16757.2) - (ODM/7677.8) - (IDM/2177.56) - (RDM/34.0948) + (WAL/27.3896)
• CDO Monitor Non-Model SDR (AA) = 0.137223 + (SPWARF/8829.01) - (DRD/20413.6) - (ODM/9556.72) - (IDM/2256.55) - (RDM/40.2751) + (WAL/26.7396)

As with the model-based approach, this SDR is then compared with the BDR at the same rating stress level (i.e., junior-most 'AAA/AA' note, excluding class X or combination notes) to determine whether the CDOM Test is passing or failing.

Similar to the CDOM non-model SDR, the non-model BDR is also generated using a regression formula that we provide once a transaction closes. We generate this formula for the non-model version using the same matrix of WAS, WAC (if applicable), and WARR values as in the model version.

In the formula-based approach, the manager or trustee simply applies the portfolio's actual WAS, WAC (if applicable), WARR (along with the adjustments to the portfolio balance to account for par gains or losses) to the regression formula to arrive at a current BDR for the transaction.

Note that this regression formula for BDRs in CDOM Non-Model is unique to each transaction, as is the input file we generate for transactions using CDOM Model (see "S&P Global Ratings' Updated Assumptions For CDO Monitor Non-Model Version," published June 21, 2019, for further details).

Why Two Versions Of CDO Monitor

We understand, from discussions with CLO market participants, that the formula version of CDO Monitor is more transparent and easier to operate.

Nevertheless, some CLO documents allow CLO managers the option to select between CDOM Model and CDOM Non-Model as a convenience. We allow CLO managers to switch from one version to the other and shift back to the original only once during the reinvestment period.

CDOM Non-Model currently can only be used to test tranches with an initial rating of 'AAA' or 'AA'. In some cases, where a CLO requires CDOM to be run for all classes of notes, the collateral manager may opt for CDOM Model in order to check the results of all classes, not just those initially rated 'AAA' or 'AA'.

We may also require that some CLOs document only the model version of CDOM instead of the above options, based on structural features in the transaction, including, for example, a CLO with a portfolio that is concentrated in lowly rated sovereigns or with a higher obligor concentration than the typical CLO pool. We have observed that the BDR and SDR calculated using the regression formula in CDOM Non-Model for such atypical CLOs diverge from the SDRs and BDRs generated by CDOM Model.

How To Handle Principal Cash From Asset Sales Or Pay Downs

If proceeds from the sale or maturity of an asset are not immediately reinvested into a replacement asset, the transaction will be left holding principal cash. For CDOM purposes, to provide an appropriate measure of the change in a portfolio's credit quality caused by the replacement of one asset (or group of assets) with another, CDOM should be run as if the "risky" asset were still in place, rather than running the portfolio with the "riskless" principal cash. If the portfolio were run in CDOM with the principal cash in lieu of the asset that was sold or paid down, the model would regard the principal cash from the asset sale/paydown as a riskless asset, which would reduce the default profile of the portfolio and improve the CDOM result.

However, because any replacement asset will have a higher risk profile than the principal cash, running CDOM with the proposed replacement asset in lieu of the principal cash will worsen the CDOM result relative to the result achieved with the principal cash included. This would be the case even if the replacement asset had higher credit quality, matured earlier, or added more diversification to the portfolio than the asset that was sold or paid down.

Thus, to appropriately assess the marginal impact of the sale of the original asset and its potential replacement with a new asset (or assets), the CDOM comparison before making a purchase should be, in our view, based on the following portfolios:

• With the original asset (or assets) still in place, in lieu of the principal cash received from the sale or amortization of the assets; and
• With the proposed replacement asset, in lieu of the original asset.

Not all cash from the sale of the asset may be reinvested. For example, the manager may sell a loan for 3 million, but only use 2 million of the proceeds to purchase a new asset. In such cases, we would expect subsequent CDOM runs to include the new asset in the portfolio and the remaining 1 million to continue to reflect the characteristics of the original asset sold.

In cases where an asset has paid down in full and left the transaction holding principal cash, we expect the manager or trustee to use the same approach as when an asset sale causes the transaction to hold principal cash. However, in this case, the asset in question will have passed its maturity date (since it has paid down and the transaction is holding the proceeds). Thus, for purposes of adding the asset back into the portfolio to run the CDOM Test, use a maturity date equal to the later of the end of the reinvestment period or one year from the current date. In addition, we expect the manager or trustee to use the same approach for principal proceeds from assets that have paid down partially.

Using the approach outlined above will require the collateral manager or trustee to track the source of principal cash held by the transaction to appropriately reflect the previous asset positions within CDOM. While this may require additional work for transactions holding principal cash from many different sources--for example, a CLO with principal proceeds received from the amortization of many individual loans--we believe the benefits of the approach, in terms of appropriately assessing the marginal impact of adding an asset to the portfolio, outweigh the incremental administrative burden.

(For treatment of defaulted assets currently held by a transaction and reinvestment of proceeds from the sale or other disposition of a defaulted asset, see the How To Treat Assets Rated Below 'CCC-' section below.)

How To Handle Paydowns To Note Balances

During the reinvestment period of a cash flow CLO, some tranches may pay down using proceeds generated from within the transaction. For example, transactions may pay a portion of their senior notes due to the failure of one or more coverage tests or the non-availability of eligible assets in the market.

CDOM typically has the initial CLO liability amounts built into both the CDOM Model and CDOM Non-Model versions. This necessitates an adjustment in some of these cases to appropriately reflect the reduced balances of the CLO notes currently outstanding. In running CDOM Model, where appropriate, enter the difference between the effective date par balance of the CLO liabilities and the current balance as an amount in the "Cash and Estimated Value of Defaulted Assets" cell of the model's "CDO Monitor Test" tab. The amount entered here serves as an offset to account for the portion of the liability balances that have paid down.

When calculating the amount to enter, the following guidelines, depending on the applicable facts, may apply:

• Calculate the amount to offset the portion of any senior non-class X tranche (including pari passu tranches) that has permanently paid down due to coverage test diversions or for other reasons, regardless of whether interest or principal cash was used to pay down the notes; and
• Do not consider any amount to offset the portion of any non-senior tranche that has paid down due to a turbo feature or for any other reason.

For instance, consider a cash flow CLO transaction that has paid down 20 million of its senior tranche due to coverage test failures, and has also paid down 5 million of the junior tranche due to a turbo feature (see table 2).

Table 2

In this example, the amount entered into the CDO Monitor as an offset to the liability paydowns should be 20 million for the reduction in balance of the class A notes outstanding:

• Start with the 20 million reduction in class A balance due to paydowns, and
• Ignore the 5 million reduction in class D balance due to the turbo feature, as this is paid using excess spread that is junior in the interest waterfall.

For CDOM Non-Model, a similar adjustment should be made such that the portfolio receives cash credit for the amount of any permanent non-class X senior note paydowns before determining the adjusted BDRs.

Adjusting The BDR Result For Changes In Portfolio Balance

Both CDOM Model and CDOM Non-Model require a similar adjustment to the generated BDR to account for changes in asset par since the CLO's effective date, to give credit for par gained or to account for par lost. This is done in CDOM Non-Model through the below BDR scaling formula, which is the same across CLO transactions:

CDO Monitor Test adjusted BDR = BDR * (OP/NP) + (NP-OP)/[NP*(1-WARR)], where:

• BDR = CDO Monitor Test BDR ('AAA' or 'AA');
• OP = Target par balance on the effective date;
• NP = Current par balance (performing collateral + principal cash + redemptions to senior-most class during reinvestment period + lower of market value and S&P Global Ratings' recovery assumptions ['AAA' or 'AA'] for non-performing collateral); and
• WARR = The applicable weighted average recovery rate (based on initial rating on highest-ranking class).

(See "S&P Global Ratings' Updated Assumptions For CDO Monitor Non-Model Version," June 21, 2019, for further details.)

For CDOM Model, users can add the change in portfolio balance in the "Cash and Estimated Value of Defaulted Assets" field on the CDO Monitor Test tab. Increases in portfolio balance will require a positive amount and will improve the CDOM results; decreases in portfolio balance will require a negative balance that will worsen the CDOM result.

How To Treat Assets Rated Below 'CCC-'

To appropriately reflect the credit quality of the performing assets in the collateral pool in CDOM pursuant to our criteria, we believe that any asset with a rating from S&P Global Ratings (as defined in the transaction documents) that falls below 'CCC-' should be excluded and treated as follows:

• Exclude the defaulted asset from the "CDO Monitor Assets" tab in CDOM Model;
• Calculate the recovery value of the asset using the lower of the S&P Global Ratings recovery assumption in the transaction documents at the rating level that corresponds to the stress being applied in CDOM (if running CDOM at the 'AAA' stress, use the 'AAA' recovery assumption; if running at 'AA', use the 'AA' recovery assumption) and the current market value of the asset; and
• Enter this amount in the "Cash and Estimated Value of Defaulted Assets" field on the CDO Monitor Test tab.

When a transaction has sold the defaulted asset or otherwise received a recovery amount from its disposition, continue to reflect the assumed recovery amount from above in the "Cash and Estimated Value of Defaulted Assets" field. Do this even after the transaction has received the actual recovery amount until the cash has been reinvested into a newly acquired asset or is used to pay down the notes.

For purposes of using CDOM to evaluate potential reinvestment of proceeds from defaulted assets, we expect the manager or trustee to employ this approach to generate the base portfolio against which to compare the model results for the portfolio with the new asset. In the CDOM run with the new asset included on the CDO Monitor Assets tab, the recovery amount from the defaulted asset should be removed from the "Cash and Estimated Value of Defaulted Assets" field to avoid double counting.

Similar to CDOM Model, for CDOM Non-Model, the recovery value of any defaulted assets should be added to the portfolio as a cash credit. That same recovery value should be added as a cash credit even after the transaction has received a recovery amount for the asset. Finally, once the recovery proceeds are reinvested, the cash credit should be removed from the portfolio to avoid double counting.

Note that some transaction documents do not require the CDOM Test to be met in order to reinvest proceeds recovered from defaulted assets.

How To Treat Asset Maturity Dates

For transactions still within their reinvestment period, we do not expect the manager or trustee to make any adjustment to asset maturity dates. Collateral managers and trustees should use the actual maturity date for each asset run through CDOM, including those that mature during the reinvestment period.

How To Handle Credit Risk Trades And Replacement Of Defaulted Assets

We monitor the performance of CLOs on an ongoing basis, where we assess the credit risk profile assuming all assets in the portfolio are performing. Certain transaction documents may allow "credit risk trades" (trades where the asset is at risk of default or impairment) not to be subject to the CDOM Test (i.e., similar to the treatment of defaulted assets). CLO market participants should be aware of such cases, which may lead to rating volatility because the CDOM Test would size the credit risk of the portfolio without such assets.

No Other Required Collateral Quality Tests

CDOM generates a BDR calculation, which encompasses changes to the WARR, WAS, and WAC of a portfolio. Hence, we have been comfortable with CLO documents that don't require meeting WAS, WAC, or WARR tests during the reinvestment period, as long as the CLO documents' definitions of such metrics used in the CDOM Test comport with our criteria.

We assess the credit risk profile of the CLO using the SDR that CDOM generates. The SDR is shaped by various parameters, including SPWARF, WAL, and diversity (by obligor, industry, and region) of the portfolio. Meanwhile, we assess the structural enhancement of the CLO using the BDR that CDOM generates. The BDR is affected by such parameters as WAS, WARR, WAC (if applicable), and par gains or losses.

In other words, we view the transaction through the lenses of the SDRs and BDRs.

Why CDOM Is Only Applicable To The Highest-Rated Class At Closing

As outlined in "Use Of CDO Monitor Simplified," April 7, 2014, in the past, the CDOM Test was routinely applied to all rated tranches of a transaction. We have observed in most instances that applying the CDOM Test to a single class of notes provides test results that are generally consistent with the results the manager would have received after applying CDOM to all the rated tranches. Going forward, for new and existing CLO transactions, collateral managers and trustees may generate and report the CDOM results only for the highest-rated class at the time we assign new issue ratings (typically the junior-most 'AAA' rated tranche), or for another tranche that we deem appropriate. If a CLO document requires the CDOM Test to be calculated for all classes of notes, the model version of CDOM should be used.

CDOM Is Not A Rating Indicator

We have frequently received questions on whether CDOM indicates likely movement in ratings (i.e., could a failing CDOM Test indicate a negative rating action, or vice versa). The simple answer to this is no. CDOM is a trading tool that enables the CLO manager to reinvest while maintaining the CLO at levels present when we first assigned the ratings. Nevertheless, a CLO has several moving parts that may cause us to take rating actions on the notes.

CDOM allows a manager to continue reinvesting even when this test is failing, with the caveat that the test should improve and eventually pass after subsequent trades. If the failure is large, a ratings committee may view it as incurable, which could lead to a rating action. Differences in sizing the credit risk may also lead to a different outcome. Furthermore, a revision to our rating framework could also lead to rating actions.

Our analysis considers five pillars to inform our rating decisions (credit quality, cash flows, legal risk, counterparty risk, and operational risk). In contrast, the CDOM Test only considers two of these pillars (credit quality and cash flows). A reduction in CDOM cushion reflects deterioration in a given transaction, but it would not necessarily trigger a rating action.

If A CLO's Parameters Change More Than Expected

The CDO Monitor matrix used for the CDOM model version and the coefficients used for CDOM Non-Model version are both based on a set of WAS, WAC (if applicable), and WARR ranges. If these parameters were to move outside the ranges in the matrix used to run CDOM, we would expect CLO market participants to reach out to us. In such cases, we would typically assess the scenario and, if applicable, provide a new CDOM input file (for CDOM Model) or BDR coefficients (for CDOM Non-Model).

Both SDRs And BDRs Should Generally Be Calculated Using The Same Model

We typically expect both SDRs and BDRs to be calculated using either the CDOM Model or CDOM Non-Model version only, unless there is an analytical rationale for using both at the same time.

Adjustments To CDO Monitor On The Effective Date.

Most CLO transactions do not purchase the full targeted amount of their collateral portfolio at closing. CLO transaction documents contain provisions directing the trustee, upon the effective date, to request a rating agency confirmation (RAC) of the ratings they had assigned as of the closing date.

As an alternative to requesting a RAC, we introduced CDOM Non-Model for effective date analysis in 2015 (see "S&P Adds Transparency To Its Effective Date Process For CLOs," April 20, 2015) with certain analytical adjustments. The analytical adjustments outlined below are only required for the effective date analysis. This is to ensure that the CDOM results are analytically consistent with our rating framework.

To analyze whether a CLO is passing its effective date analysis, two adjustments are required to the CDOM Non-Model inputs:

Calculation of WAS.   Although many CLO indentures allow adjustments to the calculated WAS to account for index floors (e.g. LIBOR, Euribor etc), the WAS used in the CDOM Non-Model effective date analysis should be calculated without any adjustment for such floors.

Calculation of principal cash.   The principal cash used in the CDOM Test BDR scaling formula should not include amounts (principal) that the transaction documents will allow to be reclassified as interest proceeds on or after the effective date.

If the analysis indicates a passing result, the CLO can assume an effective date ratings confirmation from S&P Global Ratings, provided that the following conditions are met:

• The manager or trustee certifies to us that the requirements outlined in the CLO transaction documents for the effective date process have been met, including the provision of any required accountant's reports;
• The manager or trustee certifies to us that they have run CDOM Non-Model (including the two analytical adjustments outlined above) and that the result is passing; and
• The manager or trustee provides us with an electronic copy of the portfolio used to generate the passing test results and an electronic copy of a trustee report based on audited effective date information.

CLOs that don't pass the alternative effective date process outlined above, or that elect not to use the alternative process, will be analyzed using our traditional effective date process.

For the traditional process, we require the trustee to provide us with a detailed portfolio breakdown (in an Excel spreadsheet) and an audited trustee report calculated with effective date information. After reviewing the information provided, we typically generate a full cash flow analysis to determine whether to provide the effective date RAC.

Hypothetical Stresses And Their Effects On CDOM Results

Table 3 provides a summary of potential effects on CDOM results from certain hypothetical stresses.

Table 3

CDOM Model is applicable at all rating levels (but should be run on the junior-most tranche with the highest S&P Global Ratings credit rating). In addition, atypical transactions, such as those that show less diversification than a typical CLO portfolio of broadly syndicated loans, should adopt the model approach. Additionally, the non-model approach is only applicable for standard transactions and only at the 'AAA' or 'AA' rating levels.

Looking at changes in these benchmarks before and after a trade reveals the specific factors driving a change in portfolio credit quality. These benchmarks are also comparable across standard transactions, as well as across collateral managers.

For illustration, we have run the following portfolio in CDOM non-model, applying the 'AAA' SDR coefficients:

• 150 obligors,
• WAL of 5.5 years,
• Assets equally distributed over 15 industries, and
• All assets domiciled in one jurisdiction.

We also applied the following BDR coefficients to generate the BDR results:

• Intercept = 0.10,
• WAS coefficient = 4.00, and
• 'AAA' WARR coefficient = 1.00

Finally, we adjusted certain parameters to test their effect on the BDR and SDR results. The results are summarized below (see table 4):

Table 4

Help Contact

For questions regarding CDO Monitor for your transaction, managers and trustees can e-mail CDOMonitor@spglobal.com.

Appendix

The CDOM Non-Model approach is built upon six simple portfolio benchmarks that are intended to provide insight into our view of CLO portfolio characteristics and credit quality. Each benchmark captures a component of portfolio risk taken into account by our corporate CDO criteria. The portfolio benchmarks enable CLO market participants to compare metrics across CLO portfolios, as well as assess changes within a given portfolio over time.

A description of each of the six benchmarks is provided below.

S&P Global Ratings' weighted average rating factor (SPWARF):   The SPWARF of a CLO portfolio provides an indication of the overall credit rating distribution of the portfolio weighted by each asset's par balance. The rating factor for each of the portfolio assets is determined by S&P Global Rating's credit rating (or implied rating) and the rating factor table (see appendix of "S&P Global Ratings' Updated Assumptions For CDO Monitor Non-Model Version," June 21, 2019. The SPWARF is calculated by multiplying the par balance of each collateral obligation (with a rating from S&P Global Ratings of 'CCC-' or higher) by S&P Global Ratings' rating factor, then summing the total for the portfolio, and then dividing this result by the aggregate principal balance of all of the collateral obligations included in the calculation.

Default rate dispersion (DRD):   DRD is the weighted average absolute deviation of individual asset rating factors in the portfolio relative to the SPWARF, weighted by asset balance. This is essentially a measure of how dispersed a portfolio is with respect to the asset rating factors around the SPWARF: A higher DRD benchmark means a wider dispersion of derived S&P Global Ratings credit ratings. The DRD is calculated by multiplying the par balance for each collateral obligation (with a rating from S&P Global Ratings of 'CCC-' or higher) by the absolute value of the difference between the S&P Global Ratings rating factor and the SPWARF, then summing the total for the portfolio, and then dividing this result by the aggregate principal balance of the collateral obligations included in the calculation.

Weighted average life (WAL):   This is the portfolio's average life, weighted by asset balance, based on the final maturity of each loan in the portfolio. The WAL is calculated by determining the number of years between the current date and the maturity date of each collateral obligation (with a rating from S&P Global Ratings of 'CCC-' or higher), then multiplying each obligation's principal balance by its number of years, then summing the results of all obligations in the portfolio, and then dividing this amount by the aggregate par balance of collateral obligations (rated 'CCC-' or higher).

Obligor diversity measure (ODM):   This is a measure of the "effective" number of obligors in a collateral pool, taking into account the size of each obligor as a proportion of the overall portfolio. For purposes of calculating ODM, multiple loans from a single obligor in a portfolio should be rolled up together. The ODM is calculated by determining the aggregate principal balance of the collateral obligations (with a rating from S&P Global Ratings of 'CCC-' or higher) from each obligor and its affiliates, then dividing each of these amounts by the aggregate principal balance of obligations (rated 'CCC-' or higher) from all the obligors in the portfolio, then squaring the result for each obligor, and then taking the reciprocal of the sum of these squares.

Industry diversity measure (IDM):   This is a measure of the "effective" number of industries in a collateral pool, taking into account the size of each industry as a proportion of the overall portfolio. When calculating IDM, industry categories from our corporate CDO criteria should be used. The IDM is calculated by determining the aggregate principal balance of the collateral obligations (with a rating from S&P Global Ratings of 'CCC-' or higher) within each S&P Global Ratings industry category in the portfolio, then dividing each of these amounts by the aggregate principal balance of the collateral obligations (rated 'CCC-' or higher) from all the industries in the portfolio, then squaring the result for each industry, and then taking the reciprocal of the sum of these squares.

Regional diversity measure (RDM):   This is a measure of the "effective" number of regions in a collateral pool, taking into account the size of each region as a proportion of the overall portfolio. When calculating RDM, the country and region classifications from our corporate CDO criteria should be used. The RDM is calculated by determining the aggregate principal balance of the collateral obligations (with a rating from S&P Global Ratings of 'CCC-' or higher) within each S&P Global Ratings region category, then dividing each of these amounts by the aggregate principal balance of the collateral obligations (rated 'CCC-' or higher) from all regions in the portfolio, then squaring the result for each region, and then taking the reciprocal of the sum of these squares.

Related Research

• CLO Spotlight: How To Build Your Own CDO Monitor E8 (Non-Model Version), June 21, 2019
• S&P Global Ratings' Updated Assumptions For CDO Monitor Non-Model Version, June 21, 2019
• Standard & Poor's Introduces Non-Model Version Of CDO Monitor, Dec. 8, 2014
• Use Of CDO Monitor Simplified, April 7, 2014
• S&P Adds Transparency To Its Effective Date Process For CLOs, April 20, 2015