Risk Without Reward

Transparency - One Step Forward, but Half a Step Back?

One of the most helpful initiatives to come out of the credit crisis - and Madoff in particular - is the concept of the administrator transparency report.  These documents, now in use by several administration firms and in development by a number of others, provide key metrics focused on the two primary hedge fund operational risks, existence and valuation.  Each transparency report usually indicates the percentage of assets which have been vouched to independent asset confirmations (PB, custody, counterparty statements etc.) and the percentage of the portfolio which has been independently priced by the administrator.  As an extension, some form of FAS 157 (now ASC820) presentation of the portfolio by asset pricing "level" is often included.  

These reports are helpful and unambiguously demonstrate the ability of the administrator to provide more meaningful information to the investor.  That is, we must remind everyone, the primary role of the entire administration industry...admins are meant to be, first and foremost, an independent watchdog appointed to ensure an accurate, independent and fraud free NAV calculation.

One thing that is particularly helpful in each transparency document is independent confirmation of asset existence.  While "trust but verify" is a wonderful soundbite, we think it is actually pretty unreasonable for each individual investor to turn operational due diligence into a forensic audit, with each investor looking to contact PBs and counterparties to confirm asset existence.  While this process looks wonderful in a fund of funds marketing document, in reality it is simply impossible - for every fund in the portfolio - to (i) get all managers to agree to an asset circularization, then (ii) to get all counterparties to respond (or at least respond with anything other than a no and a disclaimer), and (iii) receive from each counterparty comparable information in terms of data presentation and cut off date.  (As an aside, an asset confirmation request to Madoff Securities would have provided a wonderfully clean, check the box investor confirm).  It is, in our view, far more efficient to place the administrator in the role of intermediary; the admin can then confirm to the investor base as a whole that asset confirmations have been received independent of the asset manager and, moreover, have been reconciled back to the underlying fund accounting records.

We are therefore, one step forward - but, as in so many things, solutions are not perfect.

With respect to asset existence, we see a couple of problems.  The first is that not every document provided by a PB is necessarily unambiguous evidence of asset ownership.  A prime brokerage statement pulled from a web portal, for example, may not be an actual custody statement and can, in some cases, include "memo" assets reported by the manager.  Practices vary PB to PB, but it is worth remembering that the legacy Lancer fraud (which will finally get to court soon) relied on PB rather than custody statements to convince reputable administrators that fake assets actually existed.  In fairness, this problem is less prevalent than in the past, but investors still need to speak with each admin to ensure that they are reconciling to true, "blue border" statements, not something produced by a web reporting group separate from the bank's custody function.  The underlying issue, of course, is that junior staff at an admin may not know the difference.

Separately, are investors certain that every piece of information supporting existence of assets has been received by the administrator independently?  What happens, for example, if a counterparty statement is forwarded to the administrator by the manager?  Is this included or excluded in the "independent" verification?

Finally, many funds hold assets such as bank debt, loan originations, private equity, other hedge funds etc. which are not "custodied".  How does an administrator vouch the existence of these assets?  Would, for example, an administrator accept a binder of closing documentation and share certificates in support of a new private investment?  The snag here is that, unless the admin contacts the investee company to obtain such documents directly, there is no direct control to prevent a manager creating a fake set of transaction papers.  Moreover, private company share certificates sent to the PB for "custody" may equally not be checked for accuracy, but may be considered to be "independent" evidence of existance.

It is valuation, however, which creates the more significant issues.  The underlying point, as we have discussed before, is the shift in the administration industry from the role of "valuation" agent to "verification" agent.  When an administrator sends a transparency report, therefore, some of the prices may have been sourced by the admin, but some of them may be the manager's marks thereafter "verified" by the administrator.  Pointing out the obvious, it is usually the easy, hard to fake stuff - Level 1 exchange traded equities - which are priced by the admin, while the harder to price securities are manager marks subject to verification.

For investors, therefore, a number of questions arise.  Firstly, if prices have been verified, what are the tolerances that have been used?  Are tolerances applied at both the individual security level as well as the overall portfolio level? (investors need to ensure that a number of small differences, all in the same direction, do not become material across the entire book).  Tolerances differ by administrator, meaning that "verifications" performed by different admins may not be comparable.

Secondly, is verification passive or active?  Has, for example, the manager established a valuation policy which states that the administrator is only "responsible" for checking prices to ABC pricing feed, a designated counterparty statement, a pre-specified pricing model or quotes received from only certain brokers?  This type of "passive" process may only check prices to a pre-determined, limited selection of valuation sources, with such sources designated and potentially controlled by the manager.  "Active" verification would involve the admin applying its own valuation policies and independently searching out alternative pricing sources; there is, of course, a great difference between a process which seeks to accumulate enough evidence to justify a price, as distinct from a procedure which actively searches out evidence to disprove a valuation.  One simple starting point, for example, would be for each admin to utilize a centralized, security master pricing file to identify - and investigate - situations where other fund clients hold the same security.....but at a different price.

The final concern relates to the source of pricing data.  As above, would some administrators assume that a broker quote forwarded to them by the manager is "independent"?  As a more sophisticated example, we have encountered administrators who claimed to receive all broker quotes independently (and would, we assume, have represented as such on their transparency report).  On investigation, however, we have identified cases where managers may use a centralized "pricing" email box which receives quotes and then auto-forwards emails to the admin.  In practice, this can be very convenient but, of course, an email account hosted on a manager's server could obviously be manipulated.

The above points are not to dismiss the challenges of pricing hard to value securities and we continue to recognize - and appreciate - the hard work completed by administrators as they grapple with valuation issues across a myriad of asset types.  However, this discussion does demonstrate, as is always the case in effective due diligence, that the devil is in the detail.  Even with a transparency report, investors still need to ask detailed questions to the administrator to understand source, scope and process.

Unsurprisingly, the "step backward" in this transparency initiative is actually the risk of reduced transparency.  In some of our recent discussions with administrators, we have seen the beginning of a view that "we are now providing the transparency report, and that is the only information that will be available to investors."  Again, transparency reporting is a great initiative which we welcome - but it emphatically does not remove the need to ask follow on questions.  Indeed, simply relying on a transparency report without those additional questions may mean allocating capital with false confidence.

  

www.castlehallalternatives.com

Hedge Fund Operational Due Diligence

"Risk Without Reward" is a trademark of Entreprise Castle Hall Alternatives Inc.  All rights reserved.

Due Diligence on Managed Accounts - Essential or Irrelevant?

One of the core topics at the recent GAIM Ops Cayman conference was, unsurprisingly, managed accounts.  One speaker at the conference gave the statistic that, of an industry that may top $4 trillion in assets by 2015, 25% of this is expected to be in managed account structures.  Like them or loathe them, managed accounts are clearly here to stay.

There is broad agreement that managed accounts provide the well known benefits of TLC - transparency, liquidity and control.  We completely agree with these principles although, like many good ideas, the devil is in the details, and the details are in the execution.  Liquidity, in particular, is a very tricky topic: the same panel at GAIM Ops Cayman included a representative from a manager who raised a range of thoughtful questions in this area.  One issue was whether a manager holding hard to value securities could actually face a conflict with his fiduciary responsibility to his fund investors if he allowed a managed account to redeem during a market crisis when the flagship fund was gated.  A very good point.

From our side, we would like to raise some high level points related to operational due diligence on managed accounts.  Is operational due diligence on a managed account necessary, or simply a waste of time?

At one level, there is an argument that ops dd on a managed account does not add value - after all, the investor controls the account and, in particular, has full confidence that the assets actually exist in the investor's own PB / custody account.  If we have prevented the risk of another Madoff (assuming - ahem -  that you don't open a managed account where the assets are held in the manager's related party broker dealer!) why bother with ops dd at all? 

In our view, this position illustrates a broader disconnect as to the purpose and benefits of effective operational due diligence.  In our discussions with investors, we frequently note that the objective of operational due diligence is not solely to protect against the risk of fraud: in our view, operational due diligence is really about the broad category of business risk, rather than narrow concerns of only operations and accounting.  Even more specifically, ops dd is certainly about more that simply a focus on fraud and only fraud.

To illustrate some of our thinking, we would like to highlight five areas where we believe operational due diligence on a managed account can add value.

  

1) Reputational Risk.  We agree that a managed account provides proof that assets exist (at least for custodied assets - existence comfort can break down if the account holds participations back to the master fund for assets such as bank debt).  Moreover, the account can be structured so that a manager has no power to move assets out of the PB account.  These are powerful advantages: but, even with these benefits, a managed account does not isolate the investor from broader reputational risk. 

In the worst case scenario, let's say the manager is a fraudster, his main fund blows up, and the manager is last seen en route to his Brazilian plastic surgeon.  Yes, the managed account provides asset protection and clearly places the managed account investor in a vastly superior position - he still, we assume, has his money.  However, once the initial euphoria of avoiding disaster fades away, end investors will still ask difficult questions - why was any allocation made to the underlying manager at all?  Why was it that no-one knew the guy was a crook even as we deployed significant capital to the firm?  Did the investor rely on the managed account at the expense of asking necessary questions about the underlying manager?  In the final analysis, allocating even a cent to someone who turns out to be a crook could create pervasive, reputational damage.

In our view, this means that managed account investors still need to know a lot about what happens in the main fund.  If, for example, the manager's flagship fund is valued in an entirely different way than the managed account, the least worst outcome will be tracking error, as we note below.  At worst, you don't want to be in a position investing with a manager who turns out to have faked prices in the flagship vehicle, even if your investment was ring fenced from fraud.  Your reputation may be damaged and - even worse - with all that managed account transparency is there any risk that you might get sued too?

  

2) Business Risk.  As we noted above, we view operational due diligence to be about much more than operations: good due diligence should consider a comprehensive spectrum of business risk issues.  Whether you allocate through a managed account or via a commingled fund, many, identical questions at the asset management company remain.  Who owns the firm, and is that ownership structure stable?  How is compensation arranged?  Is there an effective alignment of interests between manager and investor, especially in the area of reinvestment of personal capital?  What is the culture of the organization?  Is there excessive staff turnover?  Is the CFO and / or COO appropriately qualified, experienced and empowered?   

More generally, managed account investors are exposed to similar risks in areas of compliance and disaster recovery.  Even if investing via a managed account, investors should still understand, for example, the quality of compliance documentation, the experience of the CCO, the outcome of recent regulatory inspections, and controls in key compliance areas such as personal trading, soft dollars, and conflicts of interest.  Equally, it would remiss of an investor not to understand a manager's disaster recovery and business interruption planning, especially if the manager is in San Francisco or Florida, facing natural hazards.

  

3) Accounting Risk.  One of the more important due diligence issues we have seen over the past year is the impact of managed accounts on existing operational infrastructure.  In more than one instance, we have seen situations where a back office has been well organized and has a successful history working with a single administrator and a set group of PBs and counterparties.  However, the manager then accepts three, four or five new managed accounts, introducing new administrator relationships and potentially new PBs.  What was once an effective back office suddenly becomes stretched and unfocused, as the team is forced to respond to ill informed questions from right and left.  This can particularly be a problem when new administrators without an understanding of the strategy (and potentially with less capable systems and people) try to implement their own accounting.

Other accounting risks include the need to handle more trade allocations, more accounts within the general ledger, and more reconciliations.  As an overarching observation, investors should recognize that $100 million though a new managed account unavoidably creates a greater operational and infrastructure burden than an additional $100 million subscription to the flagship fund.  The impact of a new managed account across multiple areas of a manager's back office should, therefore, be a key focus for due diligence.

   

4) Administration Risk.  A related point for the managed account investor is the need to understand the quality of procedures performed by your own administrator.  For your own account, will the administrator complete full service accounting, or is the admin really offering some form of month end NAV Lite?  If there is full service administration, will the admin complete accounting monthly, weekly, or daily?

The biggest issue here relates to a short cut we see far too frequently: many administrators working with managed accounts seek to populate their accounting records with a download from the PB, which is then reconciled back....to the PB.  In too many cases, the administrator accepts this entirely circular process and never gets a trade file from the manager, with the outcome that the admin never completes a three way (manager to PB to admin) reconciliation.  As such, the admin's work is limited to a representation of information already in the PB's systems.  In this circumstance, it becomes vital to understand whether the manager is at least reconciling their records to the PB because, without this check, the admin could be caught in a predictable, "garbage in, garbage out" trap.  Once more, this information can only be gathered through onsite discussion with the manager and their back office team.

   

5. Tracking Risk.  Our final point is to highlight the need to monitor tracking error between the flagship fund - the reference entity - and the managed account.  This can highlight many areas of potential operational risk, even if tracking error is anticipated.

At one level, any managed account introduces the need for effective trade allocation policies.  Does the manager allocate pre trade or post trade, and are there any wrinkes like odd lot fills, rebalancing flows and cross trades.  Thinking more suspiciously, what control prevents a manager allocating certain choice trades to the flagship fund and not your account, assuming that you have negotiated a keener fee structure? 

A separate trade issue is the need for pre trade compliance, especially as most managed accounts will introduce at least some, long-only style investment restrictions.  Issues should as leverage limits, sector concentration limits, no sin stocks, environmentally responsible investing etc. may be common in the long only world, but very few hedge fund managers have installed a Charles River or MacGregor style pre trade compliance module.  Moreover, very very few hedge fund managers are used to thinking about pre trade compliance issues before they pull the trigger on a new position.  Does the manager have something beyond an Excel spreadsheet in place to monitor trade instructions?

We have already touched on the issue of valuation above: for any hard to value security, there could easily be material differences between manager, administrator and custodian marks.  Has a valuation procedure been put in place and, for anything tricky, will it be the manager or your administrator who will price the portfolio?  Moreover, even if the administrator seeks to value (or verify value) for the portfolio, will it turn out that the manager will still have material involvement in hard to value securities?  Have you reviewed the manager's valuation policy and are you comfortable that the manager's valuation committee is actually meaningful rather than a simply a "looks good in the marketing presentation" rubber stamp?

As a final point, another source of tracking error is the difficulty of replicating more complex counterparty arrangements.  Who will be responsible for negotiating ISDAs and within what time frame?  Again, hedge fund managers are not used to thinking about allocation issues, and a key source of tracking error can be the one off trade allocated in the main fund on swap through JPM, when it turns out that an ISDA with JPM has not been negotiated for the managed account.

   

Our hope in our discussion above is to at least pose questions and illustrate the scope of due diligence issues which can arise, even in a managed account structure.  As always, the purpose of due diligence is to provide better information and make the investor more informed about risk - even if we remain entirely comfortable to accept that risk.  In this framework, we certainly see due diligence, even on a managed account, as something which can help avoid unexpected questions - let alone monetary losses - further down the line.

   

www.castlehallalternatives.com

Hedge Fund Operational Due Diligence

"Risk Without Reward" is a trademark of Entreprise Castle Hall Alternatives Inc.  All Rights Reserved. 

Can you ask an administrator what is in a portfolio?

Each year, Castle Hall completes operational due diligence on several hundred hedge funds.  As part of that process, we speak with each administrator to both understand and evaluate the quality of that admin's procedures, and also to corroborate information provided to us by each manager.  This latter part of the process is very important: it is evidently vital that the third party service provider confirm key information to validate and cross reference data provided by the manager.  

One of the questions we always ask is for the administrator to confirm the percentage of each hedge fund's portfolio held in each main category of financial instrument.  We do not ask for portfolio transparency, but do need the admin to confirm, with reasonable accuracy, the percentage of each book in equities, bonds, derivatives etc.  Clearly, if a manager indicates that 5% of the portfolio is in hard to value securities, but the administrator indicates that the percentage is actually 29%, we have a discrepancy that needs to be resolved.

We hope this is not a sign of things to come, but we recently encountered one of the "leading" administrators who stated that their new policy was to only tell investors the types of instrument in the portfolio.  That's all fine and dandy, but it doesn't really help much if the admin tells us that the portfolio contains distressed debt - we need to know whether we're talking about 5% or 50%.

To our amazement, the admin stated that they would charge what, in their words, was a "significant" fee to provide any analysis by percentage - even though (we hope!!) the administrator has a portfolio report which should list the percentages as a matter of course.  In other words, the administrator's new policy is to charge for reading out information available on their computer screen (or, at worst, available after 3 minutes with a calculator).  Moreover, the admin would charge this fee as an out of pocket expense to the fund, so the due diligence request of one investor ends up being paid by all.

As above, we hope that this is not a sign of things to come.  However, each of these individual conversations continues, most unfortunately, to undermine investors' confidence in the administration industry, and  particularly undermine the view that administrators have a commitment to fulfill a watchdog role to protect the investor.  An unwelcome precedent, to say the least.

www.castlehallalternatives.com

Hedge Fund Operational Due Diligence

"Risk Without Reward" is a trademark of Entreprise Castle Hall Alternatives Inc.  All rights reserved.  

The neverending staircase

You may have seen various versions of the optical illusion of the neverending staircase - a stairway under which, no matter how far you walk, you don't get any higher.  One of the most famous versions is the drawing by M.C. Escher, "Ascending and Descending" (see wiki article)

Why this analogy?  Well, we thought that we would open 2010 with another comment on the issue of responsibility for valuation of hedge fund portfolio securities.  Based on the various prospectuses that have recently come across our desks, "up to date" offering documents seem to be converging around a standard set of disclosures.  Unfortunately, the cumulative effect, for the investor at least, is pretty much the same as the neverending staircase.

The starting point in most PPM's (Private Placement Memorandum) is usually to state that the board of directors is responsible for managing the business of the fund, which would include, inter alia, valuation and calculation of the NAV.  Indeed, to this point, many best practice guidance papers (AIMA, IOSCO etc.) all state that the "governing body", i.e. the Board, is the entity "responsible" for valuation.

However, not so fast.

This structure creates multiple problems.  The first is obvious - corporate secretaries in the Cayman Islands, serving on the board of hundreds and hundreds of hedge funds, clearly have no capability whatsoever to value complex financial instruments.  Directors get round this minor hiccup in two ways: (i) to specify that they may rely upon information furnished to them by the Investment Manager and (ii) to delegate the calculation of the NAV to the Administrator, as we will discuss in a moment.  

While we are not lawyers ourselves, we can give a layman's view as to the degree of liability and responsibility Cayman law places on a fund director -  in simple terms, not much.  We stand to be corrected by more informed legal minds, but, in our understanding, Cayman directors have a fiduciary duty to the fund (not to the shareholders) simply to act honestly and in good faith.  The problem with this logic is that, unless a director is presented with direct evidence of fraud, every director will always be of the view that their reliance on the investment manager's security prices, for example, is undertaken honestly and in good faith.

(As an aside, as we discussed recently, one of the latest innovations from Cayman counsel is simply to put in the PPM that the Directors are not responsible for valuation at all.  Not helpful for investors, but at least one of the Cayman directorship shops tells it like it is).

So let's turn to the admin.  Usually, the prospectus starts with reassuring language that the board has delegated the calculation of the NAV to the administrator.  However, the offering document will then state that the administrator is obligated to follow the procedures outlined in the PPM related to "calculation of the net asset value".

Here the fun really starts, as this section of the PPM will usually state that it is the Investment Manager, not the administrator, who will determine the prices for any hard to value securities (such as those not listed on an exchange).  While the PPM may say that the Manager must act in "good faith", the phrase "in the sole discretion of the Investment Manager" will likely be repeated on multiple occasions.  As we have often noted, it goes without saying that if investors are to lose money due to valuation fraud, it is unlikely to be as a result of deliberate mismarking of IBM common stock - the culprits will be hard to value, Level 2 and 3 positions.

Moreover, the final paragraph of the PPM's net asset section usually ends with something along the lines of "the Board of Directors, in consultant with the Investment Manager, may permit any other method of valuation to be used which is considered to better represent fair value".  As above, the directors can, of course, accept any recommendation from the manager provided that, in doing so, they act honestly and in good faith. 

Just in case this framework was not clear enough, the administrator usually inserts a PPM disclosure stating that they are entitled to rely upon (and shall not be responsible for the accuracy of) financial data provided to them by prime broker, third party pricing services or the investment manager.  Moreover, the administrator will state that if the Board or the Investment Manager is involved in the pricing of any of the Fund's securities, the Administrator may accept such prices without any further liability.

So we come to the neverending staircase - investors hoping for independent pricing look to the board, then to the administrator...and find themselves back just where they started, with pricing responsibility - at least for anything where prices could be more easily manipulated - back with the manager.

As we look towards 2010 and the renewed growth of our industry, this issue is, in our view at least, the achilles heel of hedge fund investing.  It is nearly a decade since Lipper, Lancer and Beacon Hill, yet the core issue in these cases - managers pricing their own books without effective administrator oversight - remains unresolved.  Indeed, as we have illustrated above, the biggest change since 2002 is for fund offering documents to be far more explicit to disclose and disclaim responsibility.  This is the exact opposite of working to close this control loophole.

There are many possible solutions, starting, of course, with careful due diligence.  Looking forward, we see potential for a redefinition of service provider roles and responsibilities, realignment of fee structures with a focus on vesting and deferral, emergence of new valuation specialists in addition to traditional administrators and wider use of managed accounts.  These are just a handful of suggestions among a much longer list of tools which could be deployed.  We encourage both investors and managers to challenge the legal status quo and look for better solutions - being proactive is clearly preferable to waiting for Lipper 2.0.

www.castlehallalternatives.com

Hedge Fund Operational Due Diligence.

"Risk Without Reward" is a trademark of Entreprise Castle Hall Alternatives Inc.  All rights reserved.

At least not all administrator's think the same...

Plenty of our recent posts have focused on the declining degree of oversight and responsibility accepted by many of today's hedge fund administrators.

In light of this trend, we were very interested to see a recent paper published by a smaller admin, Crederian.  We are unfamiliar with this firm, but certainly like their philosophy, as expressed in a list of eight administrator "best practices":

Administrator’s Best Practices:

1. Maintains accounting records on a comprehensive General Ledger system.

2. Receives an independent trade blotter directly from the hedge fund.

3. Reconciles cash balances, portfolio trades and security positions against the Custodian and/or Prime Broker accounts on a daily basis.

4. Market values of securities are independently priced by a third party, with complete documentation on any security that is fair valued.

5. Subscription receipts are deposited into an “escrow” account.

6. Contributions and withdrawals are independently confirmed to the investor by the Administrator.

7. The Administrator co­signs any transfer of money from the fund’s trading account.

8. A complete month­end accounting report package that includes each general ledger account reconciled with complete documentation, be prepared, reviewed and signed­off by the Administrator.

Crederian also emphasize the need for administration work to completed by qualified chartered accountants / CPAs, which is another area where many current admins fall down.

This paper, and Crederian's positioning, suggests two potential trends going forward.  The first is whether investors will be prepared to have the same degree of confidence in hedge funds which have appointed a smaller, less well-known administrator.  Certainly, for highly complex funds, there remains an argument for a "brand name" admin, if only because of weight of resources: some funds have a dedicated 25 person plus team working solely on their behalf at one of the major administrators. Generally, however, we're often of the view that a fund may be better being a "larger fish in a smaller sea".  To be blunt, the standard of servicing from many larger administrators is so varied, and can be so poor, that we are always open to examining the service quality provided by a smaller firm.  If key metrics are met - daily accounting, sourced from the manager, daily recs, independent valuation - then investors can take strong confidence from the procedures performed, even if the admin is smaller and less well known.  Indeed, one of the biggest dangers in the industry remains investors' "check the box" mentality, irrespective of actual service quality: as one example, hiring a "top tier" admin, if all that admin does is take data from the PB, import it into their accounting system, and reconcile back to the PB, is hardly worth the basis points the investor has to pay.  This is, however, the accounting model adopted by several of the largest fund administrators for many clients.

The second trend is whether the administration industry will start to differentiate itself based on service model and service quality.  It would certainly be encouraging if we saw more firms look to position themselves based on their ability to provide meaningful, independent, "best practice" servicing.  To this point, Crederian's efforts to emphasize the need for traditional "quality" is certainly a welcome trend.  We hope more administrators follow.

www.castlehallalternatives.com

Hedge Fund Operational Due Diligence

"Risk Without Reward" is a trademark of Entreprise Castle Hall Alternatives. Inc.

Rethinking Fund Administration 1.5

In one of our last posts, we began a series of comments discussing the current state of the hedge fund administration industry, with a particular focus on valuation.  We promised that our next post would address some solutions to the "valuation dilemma", and we will return to this topic shortly.

However, we were interested to read the new version of the AIMA Guide to Sound Practices for Hedge Fund Administrators, published in September (available here).  This includes, as would be expected, commentary as to the role and responsibilities of the administrator regarding valuation.  

Before we go to our second post on solutions, we thought that it would be useful to make some comments on the AIMA paper - which holds itself out as the central benchmark of administrator "sound practice" - as an addendum to our prior post.  We hence have some comments on rethinking hedge fund administration 1.5.

The point we would like to make is to compare what AIMA now say in 2009, with what they used to say in the first version of the administrator sound practices guide, which was published back in September 2004. 

In Section 3.2, Security Pricing, the "Old Guide" stated:

Security pricing is a high-risk area for administrators, managers and prime brokers. It requires high quality data feeds, automated price validation systems and highly experienced staff. The risk of applying the wrong prices to a fund is ever present and extreme care is required to avoid this and any consequent reputational damage to the manager, administrator and prime broker.

It is also vitally important to consider the independence of the pricing process. Wherever possible, the administrator must independently price the assets of the fund. This is essential to preserve the actual and perceived independence of the NAV process and to protect the manager from accusations of bias. If the pricing environment of the fund is not rigorous, then the investors could be exposed to the risk of incorrect pricing.

However, there will be occasions when valuations of certain instruments can only be made by the manager and the issuer of the instrument, or by the manager alone. In such situations, it is important that the method of valuation is agreed with the administrator and the auditor before the fund is launched and that there is some way for both of them to check the pricing of the instruments during the life of the fund. The administrator should be able to assess the “reasonableness” of the manager’s and/or product issuer’s price.

We would completely agree that it is "vitally important to consider the independence of the pricing process".  Moreover, even for the hardest to value securities, AIMA's 2004 guidance states that the administrator (and auditor) must find "some way for them both to check the pricing of instruments during the life of the fund."

When getting to the very hardest to price securities, the Old Guide states:

In some cases, the administrator may be required to rely on prices determined by the manager, using its own pricing model for some specific complex derivatives instrument, for example. Although this is not desirable, there may be no alternative. In such cases, the administrator should:

1. ensure that the valuation procedure whereby the administrator is obliged to value the fund’s portfolio on prices based on the manager’s pricing model or formula is fully disclosed in the offering documents and the administration agreement;

2. ensure that the manager discloses both to the administrator and the fund’s auditor how the pricing model works and permits the auditor to carry out spot checks from time to time, to verify correct application of the pricing model; and

3. if possible, install the pricing model on its own system so that it can check the prices provided by the manager.

Armed with their own sources and the counterparty’s valuation, the administrator should be in a position to value the OTC position competently.

These provisions, therefore, provide specific guidance to the administrator when looking at OTC trades and, overall, create a "do as much as you possibly can" mindset.  The guide continues:

Security pricing is a highly skilled area, requiring highly trained and knowledgeable staff. It should not be done by “part-timers” nor should there be duplication of effort. Ideally, each security should be priced once for all funds holding that security at that time. Control of this function is one of the central obligations of the administrator, who should be employing common pricing procedures across all of the funds it administers.

This again suggests that there is a clear responsibility for the administrator to both disclose any manager pricing and, moreover, do all that is possible to derive independent valuations.  We particularly like the comment that "control of the [pricing] function is one of the central obligations of the administrator."  Moreover, the Old Guide makes it clear that this work should be supported by suitably skilled resources within the admin's organization.

Well, what a difference five years makes.  These are the key comments on valuation from the New Guide:

In all cases, it is important to consider the independence of the pricing process. Wherever possible, but subject in all cases to the fund’s pricing policy, the administrator should seek independent sources for the valuation of the assets of the fund. However, there will be occasions when valuations of certain instruments can only be made by the manager and the issuer of the instrument, or by the manager alone. In such situations, it is important that the method of valuation is documented, checked where a secondary source is available and reported to the pricing committee or board, periodically.

OK.  So now it is "important", not "vitally important" to consider the independence of the pricing process.  There is also a big difference between "wherever possible, the Administrator must independently price the assets of the fund" (which means that the administrator's prices would be used in the NAV) and "wherever possible, but subject in all cases to the fund's pricing policy, the administrator should seek independent sources for the valuation of the assets of the fund."  Note also that we are now in a world where there "will", rather than "may" be occasions where valuations can only be made by the manager.  We continue:

The pricing policy document should be approved by the board or governing body of the fund on a regular basis (annually or when altered).

It should be noted that while some HFAs have people, teams and service models which are capable of, and expert in, the calculation of recommended prices for individual securities, many others are not. A hedge fund manager or investor should not automatically assume that the HFA offers this service, employs this expertise or takes this responsibility.

Particular attention should be given to the governance aspects of the AIMA Guide to Sound Practices for Hedge Fund Valuation, as the observance of their principles should avoid any misunderstanding or lack of clarity around pricing and responsibility.

The Guide notes that the area of valuation has been topical and, in some cases, controversial. The role of the HFA has regularly been questioned and many investors and other industry participants have asked: “Is the HFA responsible for valuations?”

The Guide notes (and agrees) that both the AIMA and IOSCO documents suggest that it is the governing body that is ultimately responsible for the valuation of assets. In most cases, this is a board of directors or a general partner. Typically, the board will set a pricing policy.

The main thrust of the new document, therefore, is to revert attention to the pricing policy (and how many funds actually have one which really is detailed and comprehensive?) and place the administrator in a secondary, reactive role.  The guidance is much more vague, and there is certainly not the same emphasis on the "essential" need for administrators to "independently" value assets.   Moreover, there is now room for "many" administrators not to be "capable" and "expert" in the calculation of recommended prices for individual securities.  

Our biggest comment though, is one we have made before.  It remains utterly, utterly and utterly ridiculous to suggest that a Board of Directors comprised of two gentleman in the Cayman Islands with corporate secretarial backgrounds, who serve on the boards of hundreds of other hedge funds, can be "ultimately responsible for the valuation of assets" and "set a pricing policy".  For investors who allocate tens of millions - or hundreds of millions - of dollars to individual hedge funds, this comment can best be described as disingenious fallacy.  In fact, it is an insult to the intelligence of everyone concerned.

We would also point out that investors can pay the administrator millions of dollars per year for their services on some larger funds, while the average Cayman director gets $5-10,000 per fund.  That would seem to place some framework around who has the better knowledge and who is better placed to get involved. 

Perhaps the best comment, however, is the final "sound practice" guideline.

The governing body and the HFA, together with the investment manager, will then agree on how that pricing policy is to be executed. There may be specific responsibilities placed on the HFA in the execution of this policy but investors should not automatically assume a standard model.

AIMA, therefore, leaves us in exactly the same place as our own experience of current administration servicing.  While some administrators do complete thorough and robust servicing, including detailed pricing, many do not.  The message from the industry to investors, though, appears to be fairly casual - don't take anything for granted, or "automatically assume" a standard model.  In other words, caveat emptor.

The real issue here - and one that must be addressed, sooner rather than later - is that post Madoff, post financial crisis, and post the greatest stress period the hedge fund industry has ever seen, hedge fund investors have increasing, not decreasing, exposure to managers pricing their own securities.  We would have thought that all this pain would leave us with more protections - but no, we continue with all the evident risks that go along with managers marking their own books.  Moreover, everyone concerned - managers, administrators, auditors - seem to assume that this is an entirely satisfactory state of play and, indeed, is "sound practice".  

As the boy said to the naked emperor, it is not.

www.castlehallalternatives.com

Hedge Fund Operational Due Diligence 

Rethinking Fund Administration: 1

Our comments on Risk Without Reward have frequently touched on the role, responsibilities, advantages and disadvantages of third party fund administration.  As a starting point, all of Castle Hall's partners have experience in the offshore hedge fund community, the traditional "home" of administration servicing ever since the days of the 10 commandments.  Given our backgrounds, we have always been very strong proponents of the value of third party oversight over hedge fund accounting and NAV calculation.  We continue to believe that effective administration servicing is universally the hedge fund investor's best protection against loss caused by operational failure, from honest mistakes to the worst case of dishonest fraud.  

We do, however, see a number of emerging trends in the administration industry which are gradually "changing the game".  Some developments in the administration servicing model, notably better application of technology, can be strongly positive for investors.  Other trends, however, appear to be less beneficial, with changes taking place outside the scope of investor input or intervention.  It is against this background that we would like to comment, in a series of posts, on areas where investors, managers and the administrators themselves could usefully rethink the third party oversight model.

Our first topic is, unsurprisingly, valuation.  In Castle Hall's recent white paper on hedge fund operational failures, "From Manhattan to Madoff" (available here) we noted that "misvaluation of fund assets is the most significant, unresolved operational risk for hedge fund investors."

Our particular concern was "the ongoing ability of some managers to price securities themselves."  We continued that "this practice is not in any way universal, but, unfortunately, becomes progressively more prevalent as instruments become more complex and harder to price.  It goes without saying that it is precisely those instruments which are the most difficult to value which are the most susceptible to pricing fraud - it is obviously very hard to fake the price of IBM common stock."  

We then commented on what appears to be a growing "expectations gap" in the administration industry, noting:

"Today, hedge fund investors continue to expect - rightly in our view - that third party administrators paid by investors to calculate a hedge fund's net asset value should independently value security positions.  However, many of today's administrators argue that they have no responsibility for pricing and that the administrator's role is limited to that of a 'verification' rather than 'valuation' agent. Indeed, we see an increasing trend for administrators to accept manager prices for hard to value securities without further checks."

It is against this background that we were interested to see GlobeOp's settlement with a hedge fund - which the Wall Street Journal and other news sources identified as the UK based Regents Park - related to "mis-priced position values provided by a principal of the hedge fund's investment manager."

According to GlobeOp's statement on interim results issued on August 27th:

"We have settled a long-standing claim brought in arbitration

by a former hedge fund client. The claim relates to GlobeOp's

obligation to calculate and distribute the net asset value of

the fund and the impact on that of mis-priced position values 

provided by a principal of the hedge fund's investment manager. 

The arbitration will be terminated and a settlement agreement 

was signed. GlobeOp expects that the cost of this settlement

will be approximately $27 million, net of $16.5 million

of applicable income tax benefits. The amount of the 

settlement is $43.5 million, with $20 million paid at 

settlement and the remainder paid over time with interest

equal to the U.S. prime rate. The last payment will be 

February 2011. The settlement will be paid out of existing

cash resources and will be a tax-deductible business

expense during the year in which each payment is made.

GlobeOp has no further liabilities outstanding in relation

to this matter and the issue will have no impact on

current and future clients. 

Regents Park itself was named by the UK FSA back in 2006 (see FSA news release here).  As an aside, it's interesting to see how the FSA can conduct an investigation which, at the time, did not involve "any finding of misconduct or breach of the FSA's rules" which nonetheless leaves the admin on the hook for nearly $50 million - we guess that's a different issue.  

The point of our discussion here is not to take GlobeOp to task with respect to whether they did or did not perform services in accordance with their contract with the Regents Park fund.  (We have not completed due diligence on Regents Park and, therefore, cannot comment on the specific administration servicing with respect to that fund.)  In this particular case, it is evident that GlobeOp have settled claims and have made a substantial payment to resolve the issue.

Rather, our observation is more general and is directed at the administration industry as a whole: it is self evidently a really bad idea to accept prices for securities that will be used in a hedge fund's monthly NAV calculation, when those prices have been provided "by a principal of the hedge fund's investment manager".  Guys, you need to check prices, and you need to check them independently.  As we have said before, a copy of Adobe Photoshop only costs a few hundred bucks if you happen to be a hedge fund manager who is keen on changing that broker quote which will then be forwarded to the admin.

Our next post in this series will consider some of the potential solutions for the valuation dilemma - our starting point is clear and transparent disclosure so it is at least evident what the administrator has priced and what they have not.  We will also touch on the use of third party valuation agents separate from the traditional administration companies: we see potential value in bifurcating the accounting function from the valuation process.  And, of course, we'll also say that investors will have to pay more money, including to the admins themselves, if we really want top tier servicing.

However, we do have one final observation in this post.  Anyone who completes due diligence on a reasonable number of hedge funds is familiar with the administration industry's attempts to avoid responsibility for pricing.  This is evident in the legal blurb now included in many fund offering documents, together with the pages of disclaimers in typical administration agreements (we particularly like the legal provisions stipulated by those administration companies owned by prime brokers, who are usually very clear to point out that they have absolutely no responsibility for pricing, even when a hedge fund uses prices different from what the same pb uses in its own businesses.)

Above, however, we have a case where a leading administrator did need to settle a case related to allegations of mispricing.  In GlobeOp's case, the $43.5 million settlement (pre-tax) needs to be set against the fact that the firm's revenues for the six months to June 30, 2009 were only $79.2 million. (GlobeOp, as a public company, does need to disclose its financial performance - as well, of course, as the terms of the settlement itself.  Privately owned administrators do not need to provide such disclosures, which makes it easier to shield such mishaps from public view.)

Taking a step back from all of this, what really strikes us is the huge business risk much of the administration industry seems to be taking each and every time they take a price from a manager without properly checking it.  

There are two potential problems.  Firstly, all an admin company needs is one of the firm's several hundred hedge fund clients to blow up due to faked valuations, and the financial impact could potentially be threatening to the entire administration firm.  We agree that the admins have sought to protect themselves with tight contracts which disclaim responsibility, but would a court of law actually uphold all of those terms?  We would expect that there is at least some chance that a court, dependent on the facts and circumstances involved, could find that an administrator did at least have some baseline duty of care to check for valuations which were clearly incorrect.  We remain puzzled as to why admins would accept the risk that one bad apple could wound - or even kill - the entire firm.

Secondly, the administrator's greatest friend is the investor, who continues to push for third party oversight.  Many managers, however, would quite happily operate without an admin if they had the choice and didn't face investor pressure.  Administrators, therefore, need to keep the investor community convinced that their servicing is valuable and worth the money.  

The risk here is that, if there was a significant hedge fund failure caused by mispricing, the administration industry would not make friends if their response was "ah, but we did not actually have any legal obligation to check prices and were entitled to rely upon information furnished to us by the investment manager."  The worst case scenario for the administration industry as a whole would be a loss of investor confidence, should investors be faced with a fund failure where it would seem that the admin got paid but didn't do their job.  A lack of investor confidence in such an outcome - which we certainly hope will not arrive - would hardly help anyone.

Again, our next post will examine some of the potential solutions here, with a focus on short term initiatives which could better clarify the role and limits to current admin servicing.  In the long term, however, the "expectations gap" remains.  Ultimately, if you don't price the portfolio yourself, you have not calculated the NAV.  That is the 800 pound gorilla at the back of the room for all administration service providers.

www.castlehallalternatives.com

Hedge Fund Operational Due Diligence

 

How will the credit crisis impact the administration industry?

In the past two weeks, Fortis has been rescued not once, but twice.  While hardly at the top of the priority list in times of global economic meltdown, it's still an interesting question to ask what would have happened to Fortis' administration business if the bank had ceased operations.

Thinking of a different point, it's also ironic - and fortuitous - that Lehman was one of the few of the prime brokers which had not decided to create a fund admin sideline to help attract managers to the firm's PB services.

In this environment, however, both hedge fund managers and investors need to evaluate the stability and viability of fund administration entities.  There is both a possibility of direct bankruptcy and also a risk that a parent entity in financial distress could decide to close a peripheral admin business in short order.

We can immediately think of several issues:

1) If the administrator goes bankrupt, do funds have a contingency plan in place to enable them to continue operations?  Do hedge funds have copies (ideally electronic) of the administrator's accounting which could be given to a new provider?  

While audit firms in the US may not agree with us (they tend to audit the manager's accounting and largely ignore the admin), the offshore administrator's accounting forms the official books and records of a hedge fund.  If the administrator is no longer in business, does the fund have a contingency plan to recover those records so that the next NAV can be struck on a timely basis?

2) Administrators usually control cash movements for subscriptions, redemptions and fund expense payments.  

Do these cash movements pass through a cash account within the administrator?  This would, of course, expose a fund to loss if cash is sitting in an account at the time of bankruptcy.   

More discreetly, does the administrator use some form of commingled Escrow account to receive incoming subscriptions and perhaps hold cash until anti money laundering procedures have been completed?  Are hedge funds sure that these assets are properly segregated and controlled?

Separately, if the administrator is the only one with signing authority over the offshore bank account, can the manager withdraw any residual cash and process new redemption and expense payments if the signatories are no longer available?

3) If administrators complete anti money laundering and know your customer checks, does the hedge fund have access to these records to prove that AML has been performed in the event that the admin is bankrupt?

4) Does the hedge fund have access to the full shareholder's register / list of partners capital accounts to identify all investor balances in the event that the administrator is bankrupt? 

More generally, it's worth noting that many administration companies are small, independent firms which may not be well capitalized.  In an environment which sees a sharp fall in hedge fund assets through both negative performance and net redemptions, administrators' fees will also fall. Administrator financial viability is, therefore, a real issue for both managers and investors as we navigate the coming year.

www.castlehallalternatives.com

Hedge Fund Operational Due Diligence

Hedge fund assets approach $3 trillion

Following on from our recent comments on the Hedgefund.net administration survey - which showed industry assets at $2.759 trillion - the FT reports today on a further administration survey, this time by HFM Week, which puts industry assets at $2.9 trillion.  We await the first survey which puts industry assets above the $3 trillion bar.

In fact, we may well be over $3 trillion already: the Hedgefund.net survey did not include the substantial assets administered by Fortis and IFS which amount to at least $300 billion.

Aside from the question of industry growth, one other comment in the FT caught our eye.  Providing a description of hedge fund administrators, the paper commented:

"They are used by almost all European hedge funds and an increasing number of US funds, who are under pressure from investors to provide independent valuations."

Readers of this blog will be familiar with our opinion that today's hedge fund administrators increasingly provide less and less oversight over valuation (as our prior post, "should administrators stand behind their work" discussed.)

One of the biggest fallacies in the current hedge fund industry is the ongoing assumption by far too many investors that, if a recognized administrator has been appointed, that administrator will calculate the NAV and price the portfolio.  As we have discussed before, most administrators are emphatic that they are not responsible for security valuation (although exactly who is responsible for pricing remains a legal game of musical chairs).  We then have all the variations of NAV Light and, even on full service engagements, the increasing trend for administrators to accept manager marks for that portion of the portfolio which is truly hard to value.  It goes without saying that it is precisely these challenging instruments where investors have the greatest need for independent oversight and control.

We continue to believe that effective, vigilant oversight from an independent administrator remains by far the most effective - and cost effective - protection investors have against manager errors, be they honest or dishonest.  Investors need to focus their due diligence, however, to determine which administrators meet these objectives, and which do not.  After all, if the administrator does not compute an independent NAV, exactly why are investors footing the bill?

Latest administration survey

Hedgefund.net has published a new survey of fund administrators as of Q1 2008.  Usefully, the survey breaks our AUA ("assets under administration") between single manager funds and fund of funds: total single manager AUA is stated as $2.759 trillion (representing over 13,000 individual funds), while fund of funds held $1.389 trillion.

Per HFN, the largest single manager administrators are, of course, the usual suspects:

Citco:  $482bn

Citi:  $227bn

HSBC: $217bn

Goldman: $211bn

BoNY: $149bn

CACEIS: $136bn

UBS: $82bn

GlobeOP: $78bn

PFPC:  $71bn

There are some issues with these statistics, however, as there are some very notable absences.  Fortis is not mentioned; neither is State Street.  The latter have combined their IFS outsourced back office business with Investor Bank and Trust's Investors Fund Services admin group and, per a HFM Week survey as of Q3 last year (see our earlier post), would have had over $300 billion on a combined basis.  We wonder if Fortis and State Street simply declined to respond to this particular survey.

This point makes us think of several issues investors should consider when reviewing this type of information about the administration industry.

1) As above, the surveys may have respondent bias: certain firms may choose not to respond.

2) As we have often discussed, not all administration servicing is created equal.  There are enormous variations in detailed accounting and, particularly, valuation practices between funds serviced by even the same administrator.  Due diligence is always required.

  

3) Certain administrators continue to focus on the unwelcome practice of NAV Light (under this process, administrators do not maintain their own accounting records, but simply check - to varying degrees - the manager's own NAV calculation.)  We continue to believe that the main reason for the popularity of this service is that it allows funds to "check the box" and tell investors that they do have an admin, notwithstanding that there is no full administration servicing.  Some investors have policies whereby they will not invest unless an administrator is in place: NAV Light allows funds to meet this criteria, if only on paper.

On this point, investors need to become more discriminating: the point of paying for an administrator is to pay for effective, independent oversight.  Investors should not be satisfied with restricted, NAV Light procedures.  

4) Certain administrators' businesses have grown more by concentrating on outsourced back office services than fund administration servicing.  Running a manager's back office is not the same as acting as a fund administrator: in one situation, you are an outsourced provider for the manager, in the other you are an independent watchdog tasked with protecting investors and ensuring an independent, accurate NAV calculation.  

Investors should always ask their administrators a very simple question - who is your client? If the client is the manager, then it goes without saying that the admin's fees should be paid by the manager, not the investor.

5) Consolidation has brought an increase in the number of administration firms owned by prime brokers.  This raises two issues: firstly, PB owned firms, in our experience, have the widest exclusions of liability and may accept absolutely no responsibility for pricing, for example.  A recognized name does not necessarily mean more comprehensive service.

Secondly, how effective are the Chinese Walls between the admin group and the bank's sales desk?  There is an obvious and material conflict of interest if an admin group is cutting a NAV of a fund holding a book of, say, CDO positions which were sold (very lucratively) by the same bank's sales desk. If a fund generates very significant commission and sales revenue, what would happen if the administration group had a significant disagreement with the manager over pricing?  Investors need to ask specific questions as to the management of conflicts of interest.

Overall, we continue to believe in the value of effective fund administration.  That said, the evolution of the industry is causing us great concern: we see administrators aligning their businesses and services increasingly towards the manager, not the investor, and have noted a very rapid decline in administrator oversight over valuation.  These are topics we have discussed before, and, we are sure, are issues to which we will return.