01 June 2017

Amidst healthy refinancing levels of existing collateralised loan obligations, Deutsche Bank’s research team analyses why more deals are going the reset route.

Eight deals saw a refinancing of debt in April 2017. This includes an increase in resets, which are topical among CLO investors driven by larger gains to CLO equity versus refis. Deutsche Bank European CLO research from Rachit Prasad and Conor O’Toole focussed on the reset versus refi theme and pointed to opportunities in the secondary market.

What has happened?

The term ‘refi’ is broadly used interchangeably to describe both ‘resets’ and straight ‘refis’, but in this article we use them to refer to two different but related ideas. A reset is a full refinancing of all debt tranches, where the non-call and reinvestment periods of a deal are typically 'reset' back (currently to the market standard of two and four years respectively though lately a reinvestment period of five years in a couple of deals at reset has been noted by the Deutsche Bank research team), and therefore this is referred to as a 'reset' of the deal. This differs from a refinancing of only selected tranches (and all of the par of those selected tranches) which are labelled a 'refi'. In both refis and resets new tranches are issued and the money raised is used to pay off existing tranches that have been selected for the refinancing. Note that the refi or reset option rests with the equityholder and in some deals the manager. Debt holders have typically no say in the matter and as such are ‘short the option’.

Of the 57 CLO 2.0 deals that are past their non-call period, 20 have refi'ed and 15 have reset so far. However YTD end April 2017, there have been a higher number of resets (12) versus refis (9). In its monthly research for April1, Deutsche Bank’s research team noted three refis and five resets, taking the total refi/reset count to 9/12 for the year.

Resets versus refis incentives

Both refis and resets support equity distributions in the face of declining Weighted Average Spread (WAS) by lowering liability spreads. However since resets extend the life of the deal and delays bond amortisation, they keep equity distributions going for longer which a simple refi does not accomplish. Resets also enable push back on the increasingly constrained Weighted Average Life (WAL) test thresholds while maintaining higher equity distributions over the long run from extending the deal. If the WAL test threshold is left uncured, then deals will begin to pay down. Amortisation of senior bonds raises the overall cost of debt and reduces equity returns. It also shrinks the deal and therefore reduces the value of the manager fee stream. Therefore both equity holders and managers (who have the right to reset in only deals) are incentivized to push back amortisation via reset.

There are additional costs however to doing resets versus refis that come from higher arranger/legal fees or from adding Euribor floors at the time of reset in deals where they did not exist already. Indeed where Euribor floors don't exist originally, resets introduce floors and therefore reduce near term equity cashflow versus refis. There are manager specific incentives – where a manager is looking to price a new deal imminently, they may prefer a refi to prevent any cannibalisation of investor interest in their new deal. On balance however, with several deals past their non-call period already having floors, with the ability to lock in current tight spreads for longer and to push back amortisation, resets are often more favourable versus refis. While a refi does address the declining WAS issue due to lower loan spreads adequately, a reset often is more attractive because it is a case of killing multiple birds - whether WAS, WAS/WARF (Weighted Average Rating Factor) tests or WAL test constraints- with one stone.

The outlook for resets

As mentioned above, of the 35 refis and resets that have occurred, 20 have been refis and 15 resets (see figure below).

figure 18
Of the 11 cases where WAL tests were failing one month before the refi/reset occured, seven deals reset, while four deals refi'ed. Of the 24 cases where WAL tests were passing at the time, just eight deals reset while 16 deals refi'ed. This suggests a stronger incentive to reset if WAL tests are failing (see figure below).
figure 15
The research team expects that WAL test constraints and the fact that deals from the 2015 vintage and after have floors mean that reset activity should continue to increase over time and amount to c.40% of deals past the non-call period end date by the end of 2017. Since 2015 vintage deals already have floors (and therefore do not have additional costs over refis) and are likely to hit WAL test constraints in 2018, the reset probability may be higher in these deals.

What is the impact for CLO managers?

Since post reinvestment period deals don’t allow for the reinvestment of unscheduled principal proceeds in case of a failure of WAL test, these proceeds have to be used to pay down notes. Managers will likely be disincentivised to repay notes since their fees are proportional to AUMs. Equity holders too are disincentivised from allowing a repayment of notes as the weighted average cost of debt increases on paying debt down. There incentives for managers to reset are thus fourfold:

  • higher excess spread for longer means that subordinate fees/incentive fees streams are higher
  • resets result in higher AUMs for longer and therefore senior fees too are higher
  • to the extent that managers hold equity in their retention pieces, they benefit from the higher excess spread directly
  • regain the ability to trade. In some deals, managers hold the equity in their deals which would increase the probability of a reset.

1“European CLO monthly – To refi or to reset? That is the question” – Rachit Prasad, Conor O’Toole, 13 April 2017
Forecasts are based on assumptions, estimates, opinions and hypothetical models or analysis which may prove to be incorrect.

You might be interested in