划清界限

Memo to:

Oaktree Clients

From:

Howard Marks

Re:

Lines in the Sand

In my 2016 year-end review, which went only to clients, I included a discussion of the use of subscription lines by closed-end funds in areas such as private equity, real estate, distressed debt and private credit. It’s my impression that their use has become fairly pervasive in recent years, and in response to clients’ requests and market trends, Oaktree has utilized subscription lines in some of its newer funds. That year-end note prompted some interesting and spirited discussion of lines and their merit and effect. Thus I decided to write this memo on the topic for general circulation.

How Do Subscription Lines Work? As I wrote in the year-end review, subscription lines are bank loans extended to funds that enable them to use borrowed money, rather than LP capital, to make early investments or pay fees and expenses. While there is no universal description, I believe it’s safe to say in general that subscription lines:   

are limited as a percentage of the LPs’ capital commitments. (Commitments from the most creditworthy LPs earn a 90% advance rate, and commitments from lesser credits earn lower advance rates or, in some cases, zero), are secured by the LPs’ capital commitments, and generally must be repaid in the early or middle part of the fund’s life (unless extended), although terms are beginning to lengthen.

The key element is that a subscription line can substitute for LP capital, but it can’t be used to allow the fund to invest more than its committed capital. That is, a $100 million fund with a subscription line might be able to buy $50 million of assets without calling LP capital, but it still can’t invest more than $100 million in total (other than by recycling proceeds from liquidated investments). The bottom line is that essentially all subscription line financing does is defer the calling of LP capital. So the starting point for this discussion is the fact that these lines lever LP capital but do not lever funds in the sense of allowing funds to invest more than their committed capital. Fundlevel debt that allows funds to invest more than their committed capital is different from subscription lines and not my subject here.

2017 Oaktree Capital Management, L.P.

All Rights Reserved

What Are the Effects? Since a subscription line doesn’t lever a fund, its use doesn’t increase the total dollar profits that the fund will earn from investments over its lifetime (assuming the GP makes the same investments that it would have made if the fund didn’t have a line). Also, the use of a subscription line – obviously – doesn’t alter the fund’s committed or invested capital. Thus, assuming all LP capital eventually is drawn, the fund’s ratio of distributions to LP capital – either the multiple of committed capital (MOCC) or the multiple of invested capital (MOC) – isn’t improved by the use of a line. So much for what isn’t changed. The question, then, is “what is?” First the positives: 

The original purpose of subscription lines was (a) to enable GPs to make investments and pay fund fees and expenses without frequent capital calls and (b) to prevent opportunistic funds that don’t sit on large amounts of cash from missing out on attractive investments requiring quick funding. More recently, however, their use has grown for the additional reasons discussed below.



With calls for LP capital postponed, the reported Internal Rate of Return or IRR in the early years – the dollar-weighted return on LP capital – will increase substantially (assuming the early profits exceed the interest and expenses on the line).



The use of borrowed money can reduce or even eliminate the deleterious impact on early returns of the so-called “J-curve.” The J-curve results from (a) the fact that in a fund’s early years, management fees are usually charged on total committed capital, while a relatively small percentage of the capital has been put to work, and (b) the tendency of private investments to take a while to show results.



Over the course of a fund’s life, LP capital will typically be called for investments or to repay the borrowings under the subscription line. This will cause the ratio of subscription line capital employed to LP capital to decline. As a result, the fund’s IRR will retreat from its elevated early level and move down toward what it would have been if the fund hadn’t employed a subscription line. However, all other things being equal, the fund’s lifetime IRR will remain higher than it otherwise would have been, since the impact of using the line will taper off but not reverse.



Finally, any committed capital that hasn’t been called because of borrowing under the line will remain in the hands of the LPs. Thus any return the LPs earn on the uncalled capital in excess of their share of the fund’s subscription line costs will be additive to their results.

What about the negatives? 

If a fund finances investments by borrowing under a subscription line, interest and expenses will be paid that wouldn’t have been paid if LP capital had been called instead. The payment of these costs, even with interest rates below LIBOR+2%, is a permanent net negative for the fund: since the fund isn’t becoming levered, it won’t be offset by an increase

2017 Oaktree Capital Management, L.P.

2

All Rights Reserved

in dollar profits (see above). Thus it eats into the fund’s dollar lifetime gains as well as its multiple of capital. Some LPs may actually want to have their capital called and earn their preferred return. That will jibe with their expectations and preserve the historic hurdle for incentive fees. The preferred return that must be earned before the GP receives incentive fees is calculated based on how much LP capital has been called and for how long it has remained outstanding. Thus the use of a subscription line in lieu of LP capital shrinks the dollar preferred return hurdle. Lowering the hurdle can increase the GP’s probability of collecting incentive fees and cause the payment of incentive fees to the GP to begin sooner, although it will have no effect on the amount of incentive fees ultimately paid by a fund that would easily have cleared the percentage hurdle rate if it hadn’t used a line. (At the same time, however, the interest and expenses paid on the line will reduce the fund’s lifetime net dollar gains, and thus the eventual amount of incentive fees received by the GP. The interaction of these effects can be complex.) 

Less disciplined or less diligent GPs may be induced to lower the standards to which they subject investments because (a) their effective cost of capital seems so low and/or (b) they perceive an increased likelihood that the reported IRR will exceed the preferred return hurdle and thus a greater potential to earn incentive fees.



Some LPs seek to avoid so-called Unrelated Business Taxable Income (“UBTI”). Without getting into further details, suffice it to say the use of subscription lines increases the risk of UBTI to these LPs.



Since each LP’s commitment to the fund is an essential part of the bank’s collateral, the existence of a line could conceivably complicate the process of selling an LP interest in a secondary transaction, in particular if the would-be buyer is less creditworthy.



As the use of subscription lines increases, many banks are requiring greater and more intrusive information on the financial wherewithal of fund LPs to ensure the sufficiency of collateral. Some LPs are now starting to push back on providing this information, while others are expressly demanding to be excluded from borrowings, which can create an awkward dynamic among the LPs and between the LPs and GP.

Given the existence of so many pros and cons, what factors have caused the use of subscription lines to become widespread? I believe they’re these: For LPs:   

the desire for high reported IRRs, better cash management, including fewer drawdowns, and the potential to use their capital more efficiently (i.e., to use undrawn capital to make investments that may add to overall profits)

2017 Oaktree Capital Management, L.P.

3

All Rights Reserved

For GPs:    

the expectation that higher IRRs will enhance their reputations and enable them to raise more money, the potential to lower the hurdle that must be cleared before incentive fees are received, the ability to enhance reported results in a low-return world or mask otherwise-low investment returns, and defensively, a way to be competitive with other GPs who raise IRRs through the use of lines

Impact on Fund Performance Metrics The most important question in assessing fund performance is clear: Did the GP do a good job? It’s a simple question, but answering it is anything but. In particular, if a fund that used a subscription line shows a high IRR, does that confirm that the GP did a good job? Since a fund’s total dollar profits and multiple of capital aren’t improved by the use of a subscription line, the increase in IRR, while pleasant, might be thought of as illusory. Remember, as I wrote in a 2006 memo with the same title, you can’t eat IRR. My basic point in that memo was that what really matters is how much money an LP makes as a result of having committed to a fund. It’s that simple. But the deeper message was that, while valuable, neither IRR nor MOCC nor MOC – nor any other single metric – is sufficient to tell us whether the GP did a good job. There are many elements that must be taken into account, and if you hold all the others equal, one metric might be sufficient to answer the question. But the others rarely are equal. For example: 

A high IRR certainly is desirable. But that’s what a fund can show if the GP makes only one investment, with a small fraction of the fund’s committed capital, and that investment produces a substantial profit. For example, if a $100 million fund invests $1 million in something and sells it a month later for $2 million, that doubling will annualize to an IRR of roughly 400,000%. And if that’s the only investment the GP makes, that’ll be the fund’s IRR, too. But it certainly doesn’t mean the GP did a good job – I doubt the LP who committed $10 million to the fund will be happy with $10.1 million back in the end. To understand what an IRR really says about fund performance, you have to know what percentage of the capital was called and how long the GP held onto it. In short, LPs want to see their committed capital become fully invested and remain invested at solid rates of return for a long time. That’s the formula for a big gain. A high return earned on a small amount of capital for a brief period doesn’t help in that regard. High annualized IRRs on investments of less than a year can be especially misleading.



A big multiple of invested capital is good, too. But it also may be of limited significance. Let’s say the GP of that $100 million fund invests $10 million, keeps it invested for the fund’s entire ten-year life, and earns an annual return of 15% on that investment. That will result in proceeds of $40 million and thus an MOC of 4x. That’s great for the LPs . . . as far as it goes. But if that’s the only investment the GP makes, the LPs collectively will earn

2017 Oaktree Capital Management, L.P.

4

All Rights Reserved

profits of only $30 million, certainly not what they had in mind when they committed $100 million to the fund. So, again, to know if the GP did a good job, you have to know what percentage of the committed capital was invested and how long the investment was outstanding. You have to know what the GP did with the entire commitment, not just the part that was invested. 

Finally, a big multiple of committed capital sounds almost perfect. But it, too, isn’t sufficient. Let’s say all of the fund’s $100 million of committed capital is invested, and $300 million comes back to LPs, for an MOCC of 3x. That’s good, isn’t it? That depends on how long the GP kept that $100 million. If it took six years to turn $100 million into $300 million, the IRR on the fund is 20%. But if it took ten years to generate the same proceeds and MOCC, the IRR is just 11.6%. So it’s not enough to know how much capital was invested and how much was returned. We have to know how long the process took.

The answer is simple. In order to be able to assess fund performance, we have to know:    

how much capital was committed, how much capital was invested, how long it was kept invested, and how much was returned to LPs.

IRR, MOCC and MOC are all significant indicators, but none of them takes all four of those parameters into consideration. Thus no single metric is sufficient to tell us how good a job a GP did. We have to consider multiple metrics, and sometimes they will give conflicting answers. Fund A may look better on one of them and Fund B on another. So this is really just one more way in which investing isn’t subject to easy answers. Performance assessment requires consulting a variety of performance metrics; considering other factors as well, some of which are subjective (like how risky the portfolio was); and making judgments regarding the results. One fund with a higher IRR didn’t necessarily outperform another. And, provocatively, a fund that used a subscription line and came in with a high IRR may not have done as good a job – or made its LPs as much money – as one that didn’t use a line (or used a line less extensively) and reported a lower IRR. Let’s take that to its logical extreme. What if the typical race to the bottom happens at the banks, making financing available on ever-easier terms? What if we reach a point where GPs are able to obtain lines equal in size to the vast majority of their LPs’ commitments and keep the borrowings outstanding for most of the funds’ life? In that case, there will be little need for a GP to draw LP capital, and even low returns on investments could give rise to ultra-high IRRs at the fund level. The bottom line on all this is that the use of subscription lines sheds considerable doubt on the significance of IRR. And when IRR becomes suspect, anyone wanting to evaluate fund results has no choice but to put greater emphasis on the multiple of capital.

2017 Oaktree Capital Management, L.P.

5

All Rights Reserved

Bigger Questions All the above discussion is essentially mechanical, regarding matters of arithmetic. But there are other questions surrounding subscription lines that involve investment risk, and some that have bigger consequences . . . even potentially systemic. Investments are invariably viewed as safe when it is assumed that the things that should happen will happen. But I always hasten to point out that “should” isn’t the same as “will.” Let’s consider the process that’s supposed to apply with subscription line borrowings:     

The GP organizes a fund and arranges for a subscription line. LPs commit capital. The LPs put in an actual or virtual lock-box the funds they’ll need when capital is called. The GP uses borrowings under the subscription line to pay for investments, in lieu of calling LP capital. When the subscription line reaches its end, LP capital is called and the line is repaid.

That’s what’s supposed to happen. But there are ways in which actual events can deviate from that idealized progression. Most of these would be the result of negative developments in the financial markets or the larger world. 

Since the use of subscription lines results in there being fewer but larger capital calls, the magnitude of potential defaults by LPs is increased, along with the potential consequences.



Suppose the fund makes $5 million of investments against an LP’s $10 million commitment – borrowing $5 million on the line – and there’s a financial crisis (or the investments simply turn out to be big losers) and those investments decline in value to $2 million. And suppose the line comes due, the fund calls $5 million from the LP with which to repay it, and the LP – perhaps receiving simultaneous capital calls from a number of similarly affected managers – concludes it’s in its best interest (or its fiduciary duty) to NOT put up $5 million to secure investments now worth $2 million. Instead, it defaults on the capital call, depriving the fund of capital, potentially limiting the fund’s ability to repay the line and/or make further investments, and thereby possibly harming the remaining LPs. (Please note, however, that strategic defaults are an extreme hypothetical, since they would expose LPs to penalties, lawsuits and the forfeiture of their assets in the fund, in addition to the obvious reputational consequences.)



Some funds (although none of Oaktree’s) rely on subscription lines that are due on demand, rather than at the end of a stated term. What would be the effect if a large number of those lines were pulled simultaneously during a financial crisis? Or what if regulators required banks to call in their lines, even those that aren’t callable or whose terms haven’t expired?



There’s no question that the increasing use of subscription lines is altering the pattern of drawdowns and distributions. Going years without seeing much capital called could convince an LP that calls have become less likely. Suppose that, in response, rather than set aside capital equal to its commitments, the LP puts it into other investments. Although subscription lines don’t result in funds becoming levered, this kind of behavior can

2017 Oaktree Capital Management, L.P.

6

All Rights Reserved

result in the LP becoming levered (i.e., having total investments plus commitments that exceed its available capital). Now suppose a financial crisis brings large losses to fund investments in general. If the LP has made excess commitments, it could (a) suffer levered losses and (b) be forced to liquidate marketable securities in a crisis to satisfy capital calls in connection with their commitments to closed-end funds. Taken to a hopefully unrealistic extreme, could this cause LPs to become insolvent and banks to experience a wave of defaults on these lines? Market meltdowns and financial crises can increase the probability that banks will recall lines and decrease the probability that all LPs will meet the calls. If an LP has taken advantage of subscription line financing to become more than 100% committed, it might be more likely to default on the calls. If a fund has diversified commitment sources and just a couple of LPs default, the fund will probably manage just fine. But suppose many LPs default? In that circumstance it’s easy to imagine a fund being forced to sell assets during a market downturn to pay off its line and/or lacking the capital it thought it would have with which to take advantage of market opportunities. Both outcomes could be very negative for funds and LPs alike. The obligations of the LPs in a fund with a subscription line are interrelated; for example, one LP’s default on its capital commitment requires the other LPs to contribute more (up to the amount of their commitment) to repay the subscription line. Could this mean that failures by some LPs would increase the likelihood of failures by others? In the extreme, if defaults on lines are widespread, could lines become a source of significant risk to banks? In order to figure out the full impact of the use of subscription lines, one would have to know what LPs do with the uncalled capital during the period before it’s drawn by the funds. It does seem, however, that subscription lines may be adding to risk at a variety of levels. These hypothetical examples imagine financial crises, asset meltdowns and – in some cases – lessthan-conservative behavior on the part of LPs. They’re all unlikely. But are they impossible? It’s mostly during crises that weaknesses are exposed, things that are supposed to happen fail to do so, and unanticipated consequences and linkages manifest themselves. As I mentioned at the outset, some Oaktree funds have made use of subscription lines, in recognition of the advantages described above and because many of our LPs – almost all of which are sophisticated institutional investors capable of understanding how lines work and their pros and cons – have indicated that they want us to do so. However, I can report that the concerns discussed above have caused us to begin an internal process to develop guidelines intended to mitigate the risks of subscription lines while preserving their benefits. The key to financial security – individual or societal – doesn’t lie in counting on things to work in good times or on average. Rather, it consists of figuring out what can go wrong in bad times, and of only doing things that will prove survivable even if they materialize. Has anyone thought through all the implications of closed-end funds’ increasing use of subscription lines? Are they all tolerable, for the individual parties and for the financial system? I haven’t read much on this subject, but we should all be thinking about it. That’s the reason I’m writing today.

April 18, 2017

2017 Oaktree Capital Management, L.P.

7

All Rights Reserved