Bernanke's ludicrous monetary policy has forced financial companies to relever to mid-2000 levels. A recent CLO has broken the 12 month quiet period in the structured finance universe, and finally made it all too clear that bankers and asset managers have no idea what to do with all the free extra money lying around, earning nothing on the short end of the curve. Furthermore, with rampant M&A rumors every day (of which 90% are patently false) private equity must be getting nervous. As we all know, with great free money comes great irresponsibility to do really dumb things, better known as LBOs. We analyze the imminent tidal wave of going private deals, which, if Bernanke keeps ZIRP for another year as is widely expected, is just around the corner, and that the record TXU LBO of 2007 will be promptly surpassed, in both size and idiocy. Oh well, if you can't beat them, join them. We present some of the most profitable ways to play the LBO wave of 2010, and no, it does not mean tracking down Moody's Deep Shah and buying stocks 24 hours ahead of the announcement, in expectation of a 20% pop.
As CDS traders who may have been around in the long ago days of 2006 and 2007 vividly recall, back then the one and only trade in the lofty credit bubble days was to buy protection on names that were rumored to be getting acquired, playing for the LBO news spread pop. To be sure, there were several variants on the credit deterioration trade, include 3s5s steepeners, First To Default baskets, basis trades, purchases of 101 Change of Control bonds, and other such bets on credit negative/equity positive developments. Just like the LBO mania of 2007 marked the market top back then, so the same euphoria that is about to grip the markets indicates that the 12 month relevering into the biggest credit bubble of all time will be short lived. If Bernanke will stop at nothing from destroying the middle class, so be it. It does not mean that one can not profit from it (especially if the middle class itself, which has been widely warned it is now on the the endangered species list, does not care in the slightest about this development).
Bank of America's Jeffrey Rosenberg does a good and concise job of setting the backdrop for the LBO bubble v2.
They’re back. The combination of the credit market resurgence and tight spreads, attractive equity valuations and ample private equity “dry powder” create the conditions for increasing the volumes of LBOs. Whether deals will strike at the heart of the high grade market in the form of mega size transactions ($10b+) remains unclear, though the possibility clearly now exists. Unlike the most recent era, lower leverage and more prevalent change of control protections help to limit cram down losses. The IMS Health LBO illustrates the new LBO market dynamics - a $5.9B LBO funded with $3B in debt – where bank and mezzanine debt investors now augment the role of CLOs as key debt providers.
A back of the envelope analysis by Rosenberg suggests that the total pool of available LBO capital is about $70 billion. Should CLOs indeed come back, look for this number to explode.
Figure 1 below highlights our estimates of the maximum aggregate LBO volume supported by debt and equity fund raising capacity. These amounts represent only the limit on the size of LBO volumes, not our expectations of volumes in 2010. What is clear is that the return of the availability of senior debt financing is key to the ability to fund LBOs and this availability is supported by the new (relative to the earlier era) role of mezzanine debt in the "typical" LBO structure. According to these estimates of market capacity across senior, mezzanine and equity financing, expansion in senior debt financing capacity appears the constraint on the aggregate amount of LBO activity.
Note that this aggregate analysis does not describe the limits on mega size transactions – the $10B and above size transactions that garner greater attention and potential losses to cram down debt holders – as well as gains to public equity holders. That constraint remains the ability to absorb concentrated positions in a single fund. And as we describe more here, that constraint includes the new
mezzanine debt financing capacity that contributes to today’s increasing amounts of debt funding capacity for LBOs.
A summary of the various big trades that can be put on to either capitalize on PE stupidity or hedge against credit deterioration is presented below:
For those to whom the concepts presented above may be a novelty (and since there has been about 90% turnover in the trader community since 2006, that would mean pretty much anyone), here is a drill down into the nuance of each specific LBO trade. First a framework of how one should approach LBO-specific risk.
Identifying LBO Single Name Risk
After having argued for the potential for increasing volumes for LBO risk, the starting point for managing that risk is to identify names that are more likely candidates. Since definitively identifying LBO candidates is impossible, we take the other approach: exclude names in which an LBO is infeasible. By limiting the universe down to feasible LBO candidates, we create a starting point for designing hedging strategies. Moving beyond this step is both an art and a science. In the sample trading strategies below, we employ both quantitative approaches as well as bottoms up input from our team of fundamental analysts to identify this small sample of feasible (though not necessarily probable) LBO candidates.
Buying protection on a basket of LBO candidates
One way to hedge the overall market LBO risk is to buy CDS protection on a basket of LBO candidates. This hedge would clearly perform if one or more names in the basket are taken private as sub-200 bps spreads in high grade may widen to as much as 650 bps corresponding to the B-rated level in high yield. In addition, a portfolio of companies with high LBO risk would also widen in case of a rise in the overall market perception of LBO risk.
We identify a basket of five credits using our LBO screening models, see here. More precisely we select credits with the top model scores, but excluding certain sectors such as Utilities, HMO’s, and Oil & Gas companies. In addition, we impose liquidity constraints, requiring the selected credits to have a minimum level of CDS liquidity and index-eligible bonds (more than one year in maturity, fixed-coupon, at least $250mn notional). The resulting basket shown in Figure 3 r epresents a variety of sectors, with a fairly tight average spread of 82bps2 in 5Y CDS. Also, the credits are not large, averaging $2bn in incremental debt and $3bn in equity buyout amount under a typical buyout structure assumed by our LBO screen.
We put on a basket trade buying 5-year CDS protection on $5mm on each of the five names LXK, PTV, LZ, USM and HRS at an average spread of 82 bps. The target for this trade is 175 bps while we set a stop loss of 55 bps.
- One way to hedge the overall market LBO risk is to buy CDS protection on a basket of typical LBO candidates.
- This hedge would clearly perform if one or more names in the basket are taken private as sub-200 bps spreads in high grade may widen to as much as 650 bps corresponding to the B-rated level in high yield. In addition, a portfolio of companies with high LBO
- Risk would also widen in case of a rise in the overall market perception of LBO risk.
- For the basket we selected credits based on top scores in the LBO feasibility model and option volatility signals that are not inconsistent with LBO speculation, but excluding certain sectors such as Utilities, HMO’s, and Oil & Gas companies. In addition, we impose liquidity constraints, requiring the selected credits to have a minimum level of CDS liquidity and index-eligible bonds (more than one year in maturity, fixed-coupon, at least $250mn notional).
- The current average basket spread is 82 bps. The target for this trade is 175 bps while we set a stop loss of 55 bps.
Hedging LBO risk with CDS Curves
Curve trades offer a cheaper way to hedge LBO risk than the outright purchase of CDS protection. Following a successful LBO transaction, the company typically secures funding for the next three years or longer and takes out shorter-term bonds to avoid having debt maturities ahead of new debt. As a result default risk shifts out the curve, resulting in a steeper CDS curve.
To illustrate how a curve steepener trade in CDS would work consider a DV01- neutral trade in 3-year and 5-year CDS. Based on the average curve for our basket of LBO candidates this trade involves selling 3-year and buying 5-year CDS protection on the same name. Because of lower DV01 at the short maturity the trade would have roughly 30% more notional on the 3-year leg of the trade than the 5-year to be DV01 neutral. Currently such a steepener trade has positive carry of about 20bps on average for the five name basket discussed in the previous section. We estimate that such a DV01-hedged steepener would return about $13,000 per $1mn 5Y notional after an LBO announcement, assuming 570bps 3Y and 650bps 5Y spread post the announcement. The trade profits because the curve in this example steepens to 80 bps from currently 18 bps.
To the extent interest in such strategies grows along with LBO risks liquidity in this strategy may improve.
Basket Trade vs First-to-Default (FTD)
An alternative to the above proposed trade of buying protection on a basket of potential LBO names is to buy protection on a First-to-Default (FTD) contract. For example, a client could buy protection on $50mm of the 5-name basket or alternatively buy protection on $10mm of the FTD. For the above suggested basket, the basket trade costs 82bps on $50mm, resulting in an annual cost of $410,000. A $10mm FTD on the same basket roughly trades at 274bps (equivalent to 67% of the sum of spreads of 410bps), as per our estimates, resulting in an annual cost of $274,000.7 Clearly, the basket trade is more expensive than the FTD (as FTD provides protection only on the first default while the basket protection provides protection on defaults on all the names) but we discuss the trade-offs between the two trades below.
If only one name (or none) gets LBOed (or is at risk of an LBO) in our basket and the LBO risk of the other names does not increase, buying protection on $10mm FTD position is cheaper than a $50mm basket position. However, if an increase in the LBO risk affects more than one name, resulting in an increase in the risk of more than one defaults in the basket, then a $50mm basket trade will be more attractive (as it pays more if more than one name defaults) than a $10mm FTD even if the former is more expensive (Figure 8). As an FTD is a first-to-default contract and a basket trade is a contract on defaults on all the underlying names, the prices of these contracts move in accordance with the perception of default likelihood (or number of defaults) of names in the basket even if defaults do not actually happen in the short term, and therefore the intuition we are using on the perceived number of defaults in the basket holds from a mark-to-market perspective as well.
Buy Protection on the Basket & Sell Protection on FTD
If we have a basket of LBO names with low conviction, where we believe at most that there may be zero or one LBOs, from the above discussion, it is clear that the investor is better off buying protection on just an equivalent notional FTD than buying protection on all the names in the basket. However, as our basket was hand-picked based on a thorough screening and filtering of names which we think are more likely candidates for an LBO, an increase in LBO risk in the market should result in an increase in the LBO risk of more than one names in the basket. Therefore, investors could actually reduce the cost of buying protection on the basket by selling protection on an FTD.
To illustrate this strategy consider our above example. To that strategy where investor buys protection on $50mm of the 5-name basket he/she would now also sell protection on $10mm of the FTD. This results in reducing the net cost of the strategy to $136,000 from $410,000 in a naked short basket. In this example, the investor is paying 82bps on $50mm of the basket and selling protection on $10mm FTD at 274bps, resulting in a net equivalent cost of 34bps on $40mm notional short. As mentioned earlier, the key risk to this trade is - if only one name goes wider and rest all names stay tight (against our expectations) even if the overall LBO risk increases in the market.
- We have put together a basket of potential LBO names. We recommend buying protection on the CDS on each of the names in the basket on $10 million each.
- However, to reduce the cost the investor could sell protection on the FTD on the basket. As our basket was hand-picked based on our screens, we believe that there is potential for more than one name to widen with an increase in outstanding LBO risk.
- Therefore, we recommend buying protection on $50 million of the basket ($10mm each) and selling protection on $10mm FTD. The investor is essentially long protection on the second to fifth defaults in the basket.
- This trade has a negative carry of 34 bps on $40mm notional, and we expect to keep it open until this carry goes to negative 110bps. Our stop loss is a 20bps of tightening to 14bps.
A Long Correlation trade
By buying protection on all the names of a 5-name basket and selling protection on the first-to-default, the investor is essentially buying protection on second to fifth defaults in the basket. Therefore, the investor is expecting that an increase in LBO risk will likely increase the risk on all the names of the basket or, in other words, he/she is expressing a view that the risks of the names in the basket will be highly correlated. Therefore, the above trade is a way to express a long correlation view or a view that correlations will rise (among the basket names) in an LBO environment.
Another way investors could also express a long correlation view is by selling protection on an FTD and delta-hedging dynamically with single names. In this trade, the investor sells protection on an FTD of 5-names and at the same time buys protection on each of the underlying names in proportion to their deltas. The investor continues to manage the delta hedges based on the relative spread movements (or deltas) through the life of the trade. This will help in keeping the position neutral (or insensitive) to spread risk and giving exposure to the correlation risk. This position benefits from an increase in implied correlation (priced) in the basket names because of an increase in general LBO risk.
Hedging COC bond price risk with CDS
Change of Control covenants (CoC) are now common for high grade names following the last LBO cycle. These covenants offer some protection against LBO risk as they allow investors to put bonds back to the issuer typically at a price of $101 (most often conditional on a downgrade to high yield). However, given that 90% of CoC bonds are currently trading above the put price (average price of $109.6), some LBO price risk remains for CoC bond holders. Investors could buy CDS protection on the reference entity to hedge this price risk.
The $101 put price implies that the maximum loss on the bond in case of an LBO is (current price – 101). On the other hand, the maximum gain on a CDS position where the investor buys protection on the reference entity to hedge the CoC bond could be much larger than the maximum loss on the CoC bond. Therefore, the investor can buy CDS protection on a smaller notional than the bond notional.
Suppose the CDS spreads on a name jumps from 150 bps to 650 bps post-LBO. In this case the CDS hedge gain will be roughly 18 points (using a 3.2 risky BPV). To demonstrate a sample calculation of the hedge ratio, we use a CoC bond trading at $110 (the average price of the current price distribution), which has a maximum loss of 9 points, as an example. Therefore, in the event of an LBO, to be fully hedged the investor needs to buy protection on CDS on 50% (=9/18) of the notional.
Using the partial CDS hedge results in an overall long position, which means some upside from a further price appreciation on the bond in case an LBO does not materialize.
To hedge positions in bonds without change of control covenants the investor may buy matched maturity CDS protection on the same notional as the underlying bonds. This strategy works particularly well for bonds in the front end – say, three years and in - because these may be taken out as condition for the LBO financing package, as new creditors prefer not to have maturities in front of them. CDS in contrast remains outstanding and should widen significantly to reflect the shift to a more leveraged capital structure. This suggests that on average CDS underperforms front end bonds in LBO situations giving rise to positive P&L in basis package trades. Additionally, in the current market basis trades typically have positive carry with CDS spreads significantly tighter than bond spreads on the same names.
And if the above broadly shotgun approaces, and credit limited strategies are not sufficient for most readers, we present the most recently updated Goldman LBO screener (this is an excel file). The companies included represent the names most likely to be looked at actively by PE firms, and where a go private outcome would seem the highest. As such, buying the stock in a basket of the likeliest LBO candidates would be a relatively sure way to shotgun out a few quick LBO-type returns.