When it comes to looking at stocks, I’m a big believer in looking at the business from a credit perspective as a window onto equity valuation. What terms are the lenders extracting for credit and how are creditors protected? What are creditors worried about and how do they think they will get paid back? What view do the lenders (and not the company’s execs) have of the company’s liquidity? Since the imperatives of stock-owners and creditors are so different, this divergent perspective can be illuminating when analyzing distressed/semi-distressed companies in times of change.
TimkenSteel (TMST) is a perfect example of how creditors can tighten the screws, to the likely near-term detriment of equity-holders. TMST is a specialty steel company that makes steel bars used in a variety of industrial end-markets such as automotive, oil & gas, agriculture, construction, etc. As you can probably guess, anything ‘oil & gas’, or ‘industrial’-related has been under unrelenting pressure the last 1.5 years as the great commodity and US oil booms have unraveled. Since, at year-end 2014, TMST derived at least 40% of sales from oil & gas (and of that at least two-thirds related specifically to growth capex, not maintenance), it should come as no surprise that sales, margins, EBITDA, and the stock have all fallen off a cliff. EBITDA – after peaking at ~$277m in 2011 – posted $222m in 2014 then promptly cratered to negative $34m last year. This explains why the stock has done this:
I first encountered TMST mid-last year, after reading this great bearish take on the company and initiating a modest short position (around $30) upon doing further work. Back then the story was one of relatively over-valuation as much as fundamental deterioration. I covered far too early at $18 or so, and the stock kept following the steel market and US rig count lower, trading as low as <$4 on bankruptcy fears before the latest earnings release (late January) provided a bit of respite.
TMST reappeared on my watchlist after a report crossed the tape last week suggesting the company had hired PricewaterhouseCoopers to explore financing options. This struck me as a strange move, given this was clearly not the pursuing of ‘strategic alternatives’ (code for trying to sell the company), and normally financing options are handled by investment bankers, not accountancy firms. While this didn’t smell exactly of imminent bankruptcy – there are generally specialist bankruptcy restructuring advisers hired for that – something didn’t seem quite right with the announcement, so I decided to take another, deeper look at the credit.
A few things jumped out at me straight away. First off: the company renegotiated its credit agreement in December – but without putting out a clarifying press release to detail the new terms. While this is not exactly required (they DID file the required SEC disclosures), often-times – and especially when the new terms are more favorable – the company will trumpet as much to the public. In this case, unless you were an intrepid and dedicated reader of all the company’s SEC 8-K filings, you would have missed this important info until the CEO mentioned it in passing – again, in very muted fashion and without detailing terms – in the press release announcing full-year 2015 results on January 28.
In any case – the company needed to amend its credit facility because it was about to violate its interest coverage covenants on the old revolver. In return for tearing up the interest coverage (and minimum capitalization) covenants, the banks rewrote the credit facility to look like this:
- $300m asset-backed revolver, subject to a ‘borrowing base’
- LIBOR + 125-225bps rate (along with 0.4% commitment fee on unused balances)
- borrowing base governs availability – simplifies basically to 85% of eligible receivables + 70% of eligible inventory + 85% of eligible equipment, less reserves (which are at lenders’ discretion)
- company must maintain $22.5m minimum availability, stepping up to $30m minimum availability from end-June 2016 AND must have availability of $100m for AT LEAST ONE DAY prior to end-July 2016
- After July 2017, fixed charge coverage ratio covenant kicks in (FCCR must be >1:1 in the LTM period from July’17)
- Mandated max capex levels in 2016 ($45m) and 2017 ($50m)
There’s a fair bit to chew on here. But first – as of Dec 30, 2015, the company disclosed they had $200m of total debt, and $84m of total available liquidity – which, since they have $42.4m of cash, implies just $41.6m of availability on the revolver left. Now, we know that of the $200m in total debt, $18m are revenue refunding bonds unrelated to the revolver. Working backwards, then, we get an implied drawn revolver balance of $182m ($200m total debt less $18m revenue refunding bonds) and thus an implied current borrowing base of $224m (drawn revolver balance plus remaining revolver availability of $42m).
It’s important to recognize that the stated maximum availability under the revolver – $300m – is really only of secondary importance, given actual availability is governed by the borrowing base (currently $224m as discussed) – and that this in turn is a function of the company’s receivables, inventories, and equipment asset categories. As the company’s working capital and net assets contract – due to shrinking sales and lower capex – so too does the company’s liquidity.
Now let’s return to the terms governing the borrowing base calculation. Recall that the simplified calculation is: Borrowing Base = 85% x eligible receivables + 70% x eligible inventory + 85% x eligible equipment, MINUS reserves. Looking at TMST’s current situation: year end receivables were $81m (so 85% = $69m); year end inventories were $172m (so 70% = $120m), and year end equipment was ~$515m (so 85% = $438m). Sum this all up and you get $627m – and yet the current borrowing base is only $224m. What gives?
Well, we do not know the precise calculation of ‘reserves’ which are used to provide the lenders with a margin of safety; nor do we know exactly which receivables, inventories, and equipment are ‘eligible’ (though the credit agreement gives a ton of guidelines). Suffice to say the lenders here have a ton of discretion and tend to err on the conservative side, always – after all, it is their capital at risk. My own intuition suggests there are minimal reserves held against receivables, substantially more against inventories (where 70% is really not a huge discount in this environment), and much, much more against equipment (the resale value of which would be FAR lower than 85% of book in the current market).
If I were to haircut these asset categories in bankruptcy (which is ultimately what creditors to distressed companies need to contemplate), I would use an 85% rate for receivables but probably more like a 50% rate for inventories and a very low recovery assumption for any used equipment – maybe 10%. This rough methodology would yield ~$207m as a theoretical asset base for lending purposes, which is not materially off from where the current borrowing base is ($224m), suggesting banks are applying similar discounts.
Thus – the first important takeaway is that the lenders are applying fairly draconian discounts to the company’s stated asset values. This alone suggests a pretty skeptical stance from the lenders towards the company.
Unfortunately for TMST, it gets worse. As mentioned, TMST as of year-end only has ~$42m of revolver availability left. This is a big problem given a couple of the other conditions of the amended facility – namely, the minimum availability covenants. To recap, TMST must maintain minimum availability of $30m AFTER end-June, having experienced AT LEAST ONE DAY of minimum availability north of $100m.
Of course, TMST is barely in a position to meet the $30m minimum availability covenant upcoming, so the $100m one day availability covenant is simply beyond the realm. We will come to a FCF bridge shortly and explain exactly how much cash TMST needs. But cash burn aside – these minimum availability covenants, as structured, provide further insight into what the banks are thinking: they want TMST to find alternative liquidity sources, and pronto. Indeed, the ONLY reason to specifically include the one-day observation of a much larger liquidity requirement is to force the company to find another source of funds (or allow the banks to accelerate immediately if they can’t).
The Fixed Charge Coverage Ratio test that kicks in from next July is similar in this regard. In order to even approach, let alone meet, this requirement, TMST would need positive EBIT in the LTM period up to July 2017; yet EBIT was negative $108mm last year and analyst consensus is for another negative $107mm this year, and then another negative $20mm in 2017. Barring an absolutely miraculous turnaround in the oil & gas and industrial steel markets, it is wildly unlikely TMST will even sniff passing the FCCR test and thus would lose complete access to the revolver. Of course, the banks knew this when they amended the loan docs (just last December, after all) and thus the conclusion is the same – the banks are effectively demanding TMST find alternative funding in large enough size to either repay the revolver or at least to assuage the banks to accept amended terms next July.
How much cash does TMST need?
To summarize so far: banks structured the amended revolver very carefully, to a) limit their maximum exposure by applying punitive haircuts to company assets in determining the borrowing base; b) force the company to find alternative funding in the very near term to avoid tripping availability covenants; and c) ensure that these alternative funds are not ‘stop-gap’ capital and substantial enough to allow either a full bank exit, or a substantial reduction in bank risk exposure, by mid 2017.
We can now begin to think about how much cash TMST needs. While this is something of a moving target, let’s frame the discussion by looking at potential cash burn in 2016.
A rough estimate of free cash flow (FCF) is generally given by the formula FCF = EBITDA – capex – cash interest – cash taxes – cash pension contributions +/- change in working capital. See below for my estimates in each category:
- EBITDA: analyst consensus is -$28m, but the company has guided to -$15m in 1Q alone so the street is expecting sequential improvement over the year. Since rig count and industrial end-markets have done nothing but roll over further in the last couple of months, even -$28m may be a stretch. But let’s assume the lower end of consensus for now and thus EBITDA of -$20m for 2016;
- Capex: the company guided to $45m at the last earnings report; this is also the maximum amount mandated under the new credit agreement. At ~5% of sales (vs 10% 2yrs ago) this is likely bare minimum maintenance level or even below;
- Cash interest: should be $6-7m, applying loan costs to end-2015 balance, commitment fees to unused portion, and not charging much for incremental interest on additional cash burn;
- Cash taxes + cash pension payments: assume zero for both given no profits and company stated they would not pay a cash contribution towards the pension deficit ($114m) this year;
- Change in working capital: this one is a little tricky. Working capital was a massive source of cash for the company last year (+$119m) as the balance sheet shrunk and the company did everything it could to tighten its cash conversion cycle. While I do think working capital remains a source of cash (at least for a couple more quarters), I am skeptical TMST can generate anywhere close to this amount again in 2016. For one, with TMST’s rapidly declining credit quality, I don’t anticipate payable days – which tightened from 31 to 21 over the course of 2015 – rebound much. Meanwhile receivables outstanding have been cut in half – from $167m to $81m – in a year and are at the best levels the company has ever enjoyed (36 days sales in 4Q ’15). That just leaves inventories, which, like receivables, have already been tightened a lot in absolute terms ($294m -> $172m since end-2014) and in a weak environment may turn a bit slower. In any case, given a stabilizing sales picture (I assume), even further improvement in inventory days would make only marginal impact, at least in terms of affecting the company’s near-term liquidity. Eg, as the table below shows – further improvement in the cash conversion cycle (CCC) from 88 to 84 days in 1Q only releases another $5m cash at street consensus revenue estimates:
Putting it all together – TMST should burn at least ~$73m in 2016 BEFORE working capital in 2016.
But we need to think not just about total cash burn in 2016, but the progression of cashflows over the year (since the key liquidity tests under the credit agreement stipulates having $100m available at some point before July ends). A simple linear approach would suggest half, or ~$36m, will be burnt in 1H. This is probably too generous, since analyst consensus numbers are already predicated on a 2H recovery. But let’s give the company the benefit of the doubt here for a moment.
Under these assumptions, then, how much cash does TMST need before July? The below table summarizes my thought process:
To unpack it a little further – if at some point TMST needs to have $100m available on its revolver, then it needs to reduce the amount drawn on that revolver to $100m below the borrowing base. Since we know the borrowing base is currently $224m, this implies max outstanding at that point in time of $124mm. But since the company already had $182m outstanding, AND will burn another $33m in 1H 2016, that implies needed cash – to cover debt reduction and cash burn – of $91m. Two notes: this is pre-working capital, which again remains the wild-card (ie, how much benefit does the company get?). And secondly – it is highly unlikely the borrowing base remains static; indeed it will likely shrink as the asset base reduces, thereby partially or fully offsetting any working capital benefit I am not including here.
$91m may not sound like a lot, but the company’s current market cap (at $7.3/share) is only $323mm, so we are already talking ~28% of current market value at spot; of course in any dilution scenario, shares would likely be offered at a meaningful discount, thereby increasing dilution to current owners. And this doesn’t get the company through the rest of the year, nor even begin to contemplate refinancing the revolver in 2017.
This last point brings the discussion full circle. Any near-term financing designed to get the company into compliance with the pre-July availability covenants will have to also address, or provide a step towards, addressing the the upcoming Fixed Charge covenant in July 2017. From the company’s perspective, it would make no sense to issue a small amount of equity (or second lien debt), then have to come back to the market in <6 months to try to raise more capital to either refi or partially delever and convince the banks to amend the revolver (again).
I believe it is far more likely that TMST engaged PwC to try to rightsize the capital structure once and for all, in advance of the July availability test but also in such a way as to make the FCCR test next July not an overhanging issue.
Why equity and not debt? Why won’t banks ‘amend and pretend’?
At this point you may ask, ‘why would TMST issue equity and not more debt’? Well, a few reasons:
- They can’t issue more first lien debt due to negative pledges on the existing revolver (pretty standard, really, the banks don’t want more debt diluting their collateral package and currently they have a call on all company assets);
- The high yield market is closed: I shudder to think what debt capital markets desks would say if TMST – a negative EBITDA, oil & gas capex related, industrial name burning cash – called them up and wanted to do a second lien bond deal in the current environment (let alone an unsecured). Suffice to say, the window is closed and is unlikely to be open for companies like TMST at any point in the next few months;
- Equity is permanent capital and is the only solution (other than asset sales, again a non-starter in this market) that delevers the company. Deleveraging is important because it is far more likely banks agree to amend/waive the FCCR covenant next year if the amount of debt (and thus the banks’ risk) has been reduced to a reasonable level;
- The company still has a meaningful market cap ($323m) relative to the size of its enterprise value ($481m), meaning deleverage through equity sale in the market is a realistic (though not painless) option
Against this, you would have to consider management – which was buying back stock as recently as a year ago – would really have to swallow their pride to go down this road. But as they say – beggars can’t be choosers.
What is ‘reasonable’ leverage on a go-forward basis?
This of course begs the question – how much delevering would be required to assuage worried banks come next July? Well, EBITDA was $275m a few years ago and is currently running substantially negative, so it is anyone’s guess what ‘normalized’ EBITDA for a highly cyclical, boom/bust name like this is. Bloomberg consensus for 2017 gets to $51m, and then jumps to $145m in 2018. This last number seems a bit aggressive to me – especially as the company clearly benefited from a once-in-a-generation bull market in commodity capex. Let’s assume $100m is where the company could get to on a normalized, sustainable basis, at least for now.
At the moment, the company is carrying $200m of gross debt. As discussed, I think they burn at least $73m in 2016, probably more; and it is not likely they will generate any FCF in 2017 at this stage either. To that we must add $117m in pension deficit (which is effectively debt as it needs to be periodically funded) – and I wouldn’t bet against that number going up, given how interest rates have tightened so much while plan assets are probably getting killed. Putting it all together, I could envisage TMST exiting 2016 with close to $400m of adjusted gross debt, implying 4x gross leverage (and ~3.5x net) on what I think is a generous ‘normalized’ EBITDA assumption of $100m.
To be frank – this is simply too much for this kind of highly volatile, capex intensive, cyclical business, especially at this point in the cycle and currently still burning a ton of cash – even if you did believe the company could get back to $100m EBITDA in 2.5 years. 2x gross leverage (and <1x net) would be more palatable (thinking like a bank) and would help assuage doubtful bankers to amend/waive covenants next July. But again – that would dictate a very chunky equity offering (around $200m – assuming such size could get done). Without such a move I just don’t see how the company can extract much leeway from a banking group that seems pretty fixed in ‘self-preservation’ mode (and not in ‘save TimkenSteel’ mode).
Summary: TMST is between a rock and a hard place
I don’t mean to be too negative: a bankruptcy filing, while definitely a possibility, is not my base case outcome at this stage. But I do envisage a substantial (30-40% of market cap? or more?) dilution in the very near-term – whether by public offering or third-party allotment to key stakeholders, or both, and likely at a meaningful discount to where we currently trade. Such a move will be deeply unpalatable to management who were buying back stock as recently as a year ago and remain key backers of the company – but without it, I simply don’t see how they will be able to jump through the hoops assigned for them by their increasingly skeptical banks. In this scenario, I would not be surprised to see TMST retest the recent lows or at least get close, as shareholders wake up to the dilutive implications of right-sizing this currently sinking ship.
Disclosure: short TMST. All data is from company filings, Bloomberg consensus and/or my estimates.
Extra Disclosure: TMST trades like Conor McGregor fights – violently and without mercy. (up and down). Extra special ‘big boy’ risk is involved if you traffic in this name.