Is it time to sell hybrids?

The back-story...

I try to put my money where my mouth is. A couple of months prior to the Federal Election, we shifted more than $100m into franked ASX hybrids for several reasons, including: 

1. The risk of cash refunds on franking credits being eliminated by Labor was, on our numbers, already fully priced; 
2. We thought ScoMo was more likely than not to win the election; 
3. Even if he lost, we judged that if Labor won, the Senate would block any attempts to remove cash refunds on franking credits for most Australians; 
4. If Labor won and for some reason the Senate did agree to remove cash refunds on franking credits, given over 90% of investors pay tax and can still use franking credits (just not the cash refunds), we believed the market should continue to trade on a fully franked basis. This is because those who could no longer claim cash refunds would presumably sell to those who could utilise franking credits meaning the net result would be hybrids would be held by investors who can price the franking benefits (super fund providers like Netwealth were also finding ways to allow SMSFs to claim cash refunds in this eventuality by effectively pooling their credits with other members who could exploit them); and
5. Taking all of the above into account, the ASX hybrid sector looked quite cheap. 

We argued consistently that, as a minimum, credit spreads on five year major bank hybrids listed on the ASX should compress from their ~356 basis point level above the quarterly bank bill swap rate (BBSW) to around ~300 basis points over---delivering capital gains of some 2.8 percentage points on top of the income they were paying. Of course, the rest is history: ScoMo won and the ASX hybrid market went on the mother-of-all rallies with credit spreads on five year major bank hybrids contracting to 278 basis points today, furnishing capital gains of 3.7 percentage points on top of their income. 

Hunting housing and hybrid bears

Indeed, hunting hybrid haters who were aggressively pushing alternatives like high yield Listed Investment Trusts, and actively advising mums and dads to exit their franked hybrids before the election, has been almost as much fun as hunting housing bears. 

Notwithstanding the hysterical and ultimately completely erroneous "franking fracas", the ASX hybrid market was the second best performing asset-class in 2018 behind government bonds with a 4.9 per cent total return (a year in which equities were smashed), and over the last 12 months have returned a staggering 9.07 per cent accounting for the post-election rally.

As an aside, I consider myself neither a housing bull nor bear. We accurately forecast the strong house price growth between 2013 and 2017. When house prices were still rising in April 2017, we were the first mainstream analysts to declare the property boom over and predicted a 10% peak-to-trough fall in national prices, which would have been the largest on record. In April 2019 while prices were still declining, we said the correction was over and national prices would rise by 5% to 10% in the 12 months following the second RBA rate cut. 

Well, national prices did indeed correct by 10.7% on a peak-to-trough basis between 2017 and 2019 according to CoreLogic. They further flat-lined one month after our latest call in May 2019, and started increasing again over June (in Sydney and Melbourne) and July (nationally). 

During the boom (bust), I was characterised as a bull (bear). I don't care for labels: I would rather be right, and make money, than a perma-bear who has been consistently wrong. It is frankly embarrassing that the media give the latter cohort so much air-time. I guess sensationalism sells.

Should we sell hybrids?

So, to answer, the first question, Should we be selling hybrids, my answer is: Hell yes, if you shifted more than $100m into franked hybrids prior to the election (to be clear this is not financial advice---I am explaining our own decision-making). 

As hybrid spreads have crunched in from 365 basis points to 278 basis points above BBSW, and capital values have surged 3.7 percentage points, we have been taking profits almost every day. But, this is a very different question to the more fundamental one of whether we want any hybrids exposure in our portfolios. My answer to the latter question is emphatically "yes" right now, which is why we still have them.

Getting hybrid spreads right and wrong...

This is where life gets a little more complex. First, if someone is screaming that hybrids should be sold, I would question whether they were also vocally negative on hybrids prior to the election. It is possible they are pushing an "axe", like trying to sell private debt and/or high yield alternatives, or, say, over-the-counter (OTC) bonds because that is their stock in trade (ie, they do not offer hybrids). I am not sure---you should check.

Turning to valuation fundamentals, this is how we think about the hybrid market right now. We start with historical "technicals" or trading behaviours. This is where I often see pundits pushing people to exit hybrids get their numbers totally wrong... 

Not many folks actually measure or track hybrid credit spreads carefully. For years we have calculated our own internal hybrid spread curves, which is why I can tell you exactly where five year major bank hybrid spreads are trading right now. I see other pundits quote hybrid spread levels, but they are actually combing all maturities (eg, 1, 2, 3, 4, 5, 6 and 7 year hybrids). This results in them comparing, say, five year major bank Tier 2 subordinated bond spreads with the wrong (ie, non-5-year) hybrid spreads. 

The right numbers for these two securities currently are around 189 basis points over BBSW in the case of Tier 2 bonds and 278 basis points over BBSW in respect of five year major bank hybrids. By way of context, five year major bank senior bond spreads are trading at about 60 basis points over BBSW.

This is interesting because Tier 2 bonds and Additional Tier 1 (AT1) hybrids share similar, but not identical, risks. For the major banks, AT1 hybrids are rated right below the cusp of "investment-grade", which is BBB to AAA, or at BB+, while Tier 2 bonds are two notches higher at BBB, which is in turn five notches below AA- rated senior bond ratings.

(To recap, in the typical bank capital stack you start with ordinary equity, followed by AT1 hybrids, then Tier 2 subordinated bonds, above which rank senior unsecured bonds, then deposits, and finally secured senior, or covered, bonds.)

Tier 2 bonds and hybrids share similar risks

APRA can convert both Tier 2 bonds and AT1 hybrids into equity if a bank is declared to have become "non-viable" at the same time (subject to certain conditions). (In our modelling we assume that if this ever happens, the securities are completely written off.)

AT1 hybrids get automatically converted into equity if a bank's common equity Tier 1 capital ratio falls from circa 10.5% presently to 5.125%, and APRA also has the ability to restrict a bank paying out profits to both equity and hybrids as the bank's CET1 ratio approaches 5.125%. (These latter two risks do not apply to Tier 2 bonds.)

In practice, most bankers think that APRA will convert Tier 2 and hybrids into equity well before the CET1 ratio gets near 5.125% (given the huge losses a bank would have to incur to crush its equity down to this level). Most bankers also believe that Tier 2 bonds and hybrids will be converted at the same time.

Interestingly, APRA recently appeared to confirm as much when it wrote to the banks in July and stated that it would convert Tier 2 bonds into equity (or write them off) well before a bank was ever at risk of becoming insolvent. This means APRA is treating Tier 2 debt as going concern capital that can be switched into equity while a bank is still alive, but approaching death's door. This is distinct to the way the market treats Tier 2, which around the world is meant to be gone concern capital that is only converted into equity once a bank is dead as a door knob. This is important because it increases the probability that APRA will convert Tier 2 and AT1 hybrids into equity at the same time given for going concern purposes (AT1 hybrids are officially part of the going concern Tier 1 capital buffer).

While there are different risks across these securities, the key existential credit hazard, which is being bailed-into equity by APRA, are, therefore, very similar (technically, AT1 hybrids have to be bailed in first then followed by Tier 2, although the decision to do so can occur simultaneously).

The $90 billion question is whether the current 89 basis point per annum spread differential between AT1 hybrids and Tier 2 bonds correct? I will answer this later by undertaking a first-principles bottom-up valuation. Certainly if you accept the logic that Tier 2 and AT1 hybrids (increasingly) share similar risks, it suggests that either AT1 hybrid spreads should be trading closer to 189 basis points or Tier 2 should be trading closer to 278 basis points...

Historical technicals

One thing we can say with confidence is that both Tier 2 bonds and AT1 hybrids are currently paying spreads that are multiples their pre-GFC levels. Back in 2007 Tier 2 was trading below 50 basis points above BBSW while AT1 hybrids were trading at circa 125 basis points over. But the securities back then were a little different. In the case of AT1 hybrids, they did convert into preferred equity (not ordinary equity) if a bank's capital ratio declined materially, but neither Tier 2 nor AT1 hybrids allowed APRA to unilaterally bail them into ordinary equity or write them off if equity conversion is not possible.

These new so-called "Basel 3" terms were introduced in 2013. Since that time, AT1 hybrid spreads have traded as tight as 240 basis points above BBSW in July 2014. At that juncture, Tier 2 bond spreads were around 157 basis points, though they did get irrationally tight in 2018 with five year major bank Tier 2 spreads moving to as low as 140 basis points in February last year (compared to their current 189 basis point level).

Of course, the banks look very different to the institutions they were back in 2007 or even in 2014. In particular, their credit risks are much lower. CBA's common equity Tier 1 capital ratio in 2007 was just 4.7% compared to around 10.5% today. Indeed, the major banks have generally halved their risk-weighted leverage and massively de-risked their business models more widely, getting rid of most non-core exposures, including funds management, financial planning, life insurance, proprietary trading, and many offshore businesses. They are radically more risk-averse today, which is good for debt but bad for equity.

So while on the one hand the Tier 2 and AT1 hybrid terms in 2019 are less creditor friendly than they were in 2007, the balance-sheet risks to which creditors are exposed have also reduced materially. 

On a technicals basis, we have felt that there is no reason why AT1 hybrid spreads could not converge towards their mid 2014 levels. One key driver of this is the ongoing reductions in the RBA's cash rate and the huge search-for-yield bid that this has induced, which is only likely to intensify over time as credit spreads tighten. 

We also know that the current ASX supply pipeline for hybrids looks very benign with no major bank hybrid maturities this year, although we do expect at least one ASX deal and/or an OTC transaction. ASIC's new Target Market Determination rules could also encourage the banks to do more OTC transactions targeting wholesale investors, which might suck supply off the ASX as we have seen Tier 2 bonds shift from the ASX into the OTC domain in recent years.

Bottom-up valuation levels...

To really invest intelligently in this sector, one needs to forecast the probability of default on both Tier 2 bonds and AT1 hybrids, investors' likely recovery rates, estimate the minimum expected loss and then compute what credit spread above BBSW this would require one to earn to compensate us for these risks before any liquidity considerations. 

In this context, we take a very conservative approach that ordinarily involves a relatively high probability of default (PD) with a subsequent 100% write-off. (These PDs actually change all day, every day, as risk evolves.) In the case of CBA, if we estimate an 8.25% probability of a Tier 2 bond or hybrid being converted into equity, and then assume we have a 0% recovery (ie, are written-off), our minimum required return is about 172 basis points over BBSW. As another illustration taking NAB, we have an 8.5% probability of default right now with 0% recovery, which gives us a similar minimum required spread of 177 basis points. Now this is before any extra risk premium one might require for the relative illiquidity of Tier 2 and AT1 hybrids vis-a-vis government bonds.

Tier 2 supply tsunami...

One reason to favour AT1 hybrids over Tier 2 bonds is that the latter are going to be subject to a supply tsunami that the market is not yet pricing properly. On the banks' and our estimates, the four majors will have to issue $80bn to $90bn of Tier 2 bonds over the next 4.5 years, which will require the global supply of Tier 2 to expand by about 37%. This is a huge challenge, and there will be a lot of this product issued in Aussie dollars, which one major bank says will be the "dominant" currency. In contrast, AT1 hybrid issuance should be business as usual. Given this supply shock is focussed on Tier 2 (not AT1), and AT1 is currently paying 89 basis points annually in extra income, I prefer the latter sector for the time being.

We are also forecasting that both Tier 2 sub debt and AT1 hybrids will receive credit rating upgrades from S&P in the next year, which will impact AT1 more than Tier 2 because the former will be lifted from BB+ to BBB- into the all-important investment-grade band that will enable greater institutional allocations.

Hybrids are complex securities and hard for retail punters to value. But this is all relative. Hybrids are much simpler than equities to price and value given their pre-determined rate of return coupled with the abovementioned event risks. This is borne out in their volatility and maximum draw-downs. Over the last decade major bank hybrid volatility has been less than one-quarter of that of major bank equities' volatility. And during the GFC their maximum total return losses were less than half that of major bank stocks.

Please read the disclaimer below and note this is not financial advice, just our views on what we are doing in our portfolios right now.