Astrea VI PE bond: the merits of a de-risking structure

ASTREA VI, the latest issuance of private equity-backed bonds by Azalea Asset Management, carries a lower interest rate than earlier issuances. But the retail tranche with an annual interest rate of 3 per cent is still set to be very well subscribed.

The offer closes on March 16 at noon, and trading on the Singapore Exchange begins on March 19. The entire issue size is US$643 million, of which S$250 million is available for public application.

Azalea's provenance - it is an indirect subsidiary of Temasek Holdings - likely plays a part in enhancing investor confidence.

But beyond the Temasek parentage, the key reason why investors would be drawn to the Class A-1 bonds of Astrea VI is that 3 per cent is arguably a good return relative to the bond's risk. And this is thanks to the bond's structure.

From the time of issuance, the structure single-mindedly sets out to de-risk itself to ensure that investors in the lowest risk Class A-1 bonds will be paid their interest and principal.

There are, of course, risks particularly in the wake of a world still grappling with economic fallout from Covid-19.

Here are the comparative rates of other low-risk options: Ten-year Singapore Government Bonds yield 1.52 per cent; the latest issue of Singapore Savings Bond is quoted at an annual 1.15 per cent return over 10 years. Reits offer the possibility of higher yields, at the moment an average of about 6 per cent. But they come with equity-like volatility and the possibility of capital loss. As for corporate bonds, with a minimum investment of about S$200,000 they are not as accessible.

First off, you have to be clear that this bond is targeted at fixed income investors. It offers no growth potential despite the fact that its collateral comprises a portfolio of potentially high-return private equity funds. It offers a regular income and eventually your principal back. There is no guarantee, of course.

Here are some ways the sponsors have designed a systematic de-risking of the bond:

• Prescribed priority of payments. Cash flows from distributions from the underlying PE funds flow through a set sequence or "waterfall". First is to pay key expenses; then bond interest to bondholders every six months, and the principal at the expected call date in March 2026. Every six months, an amount must be paid into reserves. Payment to sponsor is last in the chain.
• Loan to value (LTV) cap. The LTV cap is set at 50 per cent. The priority of payments helps to ensure the cap is not breached. If the LTV is breached, any cash that is due to the sponsor is diverted into the reserve account to lower the LTV ratio. This feature was tested in Astrea IV and V in mid-2020 when the sponsors of the two issuances thought it prudent to direct cash which would have been due to them under the priority of payments, into the reserve accounts to reduce the net debt positions. Astrea VI starts with a LTV of 44 per cent.
• Sponsor sharing. A performance threshold is set for Class A-1 bonds. If this threshold is met, 50 per cent of the cash flow that is due to the sponsor according to the priority of payments, will be allocated to the reserve account. This enables a faster build-up of the reserve account to redeem the Class A-1 bonds on the scheduled call date. Meeting the threshold also triggers the payment of an additional 0.5 per cent of the bond principal at redemption.
• Alignment of interest. To ensure an alignment of interest, the sponsor holds the entire equity stake worth US$841 million or about 56 per cent of the net asset value.
• Fixed portfolio. The underlying portfolio is static, which means the manager is not allowed to use cash flows to take additional risks such as invest in new funds
• Seasoned portfolio. The underlying portfolio valued at about US$1.45 billion comprises 35 funds with exposure to 802 investee companies. The weighted average age is 5.8 years. S&P in its pre-sale credit report said that compared to a portfolio of less seasoned PE funds, the funds "may exhibit more net positive cash flows during the transaction's 10-year legal term because they are generally at points on their expected J-curves where most of the capital calls have occurred".
• Counterparty risks. The structure does incur counterparty risk arising from the credit facility provider which is DBS, as well as hedge counterparties. There are provisions, however, that if a counterparty is to be replaced, if its rating for example, falls below a certain level, the replacement must not cause a downgrade of the prevailing rating of the senior outstanding class of Astrea bonds.

As there is potential currency mismatch - Class A-1 bonds' interest and principal are to be paid in Sing dollars and underlying investments are in US dollar and euro - the issuer enters into hedge agreements with a number of counterparties.

Fitch notes that there is a "flip clause" in the priority of payments, which places any termination payments due to a hedge counterparty that is in default in a junior position in the chain of priority of payments. This is to mitigate the impact caused by the default or non-performance of the counterparty. "In case the issuer does not pay a hedge counterparty, the transaction documents include a 'non-petition' clause that prevents the counterparty from causing the issuer to file for bankruptcy."

• Market/investment risk. Cash distributions by PE funds are typically volatile and irregular. The pandemic and market stress could exacerbate this volatility, as managers may delay exits. This raises the risk that Astrea VI could receive less distributions.

The issuer, however, has lined up a credit facility that can be tapped for taxes and expenses, interest payments and capital calls. In addition, Fitch says the legal maturity date of 10 years "could be supportive in weathering a potential market downturn".

The prospectus also describes the use of hypothetical scenarios of increasing severity to stress test the portfolio. An independent consultant Bella Research Group was also commissioned to stress test the portfolio. In both cases, there was no default of Class A-1 bonds. In the Q&A session with investors last week, the manager cautioned that the scenarios were hypothetical and backward looking.

The experience of Astrea IV and V in 2020 could give an idea of the bonds' resilience. In June 2020, both issuances suffered fair value losses due to Covid-19. By December 2020, they recovered to pre-Covid levels. In a note to the Q&A with investors, the issuer said both Astrea IV and V portfolios generated sufficient cash flows to fulfil all bond obligations, and their credit facilities were not utilised.