Historically, the financing of investment property in the UK has been concerned with one key factor, the open market value of the underlying property asset. However, investors and lenders are increasingly questioning the prominence given to that factor in appraising whether and to what extent an investment is bankable.
In general, property investors looking for sources of finance look either to commercial banks or the capital markets. The advantages of the commercial banks is an inherent flexibility but the disadvantage is relatively short term financing. Whilst capital markets offer opportunities for long-term financing, does the price of inflexibility outweigh the advantage?
The purpose of this paper is to focus on these two areas.
Value is dead, long live value?
What is value?
How much is a property worth to an investor? Ignoring any "hope" value attributable to re-letting, redevelopment or the like, and taking as read that it is let on an institutionally acceptable, full repairing and insuring lease(s), the value that an investor would attribute to the property is determined by factors such as the amount of rent receivable per annum, the quality of the tenant(s), the unexpired term of the occupational lease(s) and whether they are subject to any break clauses. Typically in the UK, institutionally acceptable leases are for terms of 10 to 15 years or even, in the case of retail parks, up to 25 years, with upward only rent reviews occurring every 5 years. If the property is unlet then its value is far more difficult to ascertain. What are the chances of re-letting it, what sort of tenant would take a lease, for what length of term and at what rent? Would a prospective tenant require an option to break the lease, which would reduce the period for which the rent receivable would be certain? With a fully let property and tenants of good quality, the investor is really looking at the value of the future receivables rather than necessarily the value of the bricks and mortar or to any hope value. Of course, a developer, or an investor interested in improving, re-letting and otherwise "working up" the property, may be willing to pay an enhanced price for the same property, but the additional value is uncertain.
Why is value relevant?
Lending on loan to value seems natural and safe. Surely, if a lender will agree to lend only, say, 60 percent of the value of a property, even if the market declines by 40 percent he must be sure of getting all of his money back. Reducing an asset to a single value, which a professional can ascertain for you, seems to be a safe way of lending. However, what happens when the market falls because nobody wants to, or can afford to, buy? Lending on value alone, or value primarily, in this manner may well become false comfort.
There are two principal financial covenants used in property financing. The first is a loan to value covenant and the second a short-term income to debt service covenant.
The first ratio compares (a) the outstanding amount of the principal remaining to be repaid to (b) the market value of the assets charged at the time the covenant is tested. Loan to value used to be the primary covenant but, as referred to above, value is a somewhat elusive concept. A number of property companies breached their loan to value covenants when the property market collapsed ten years ago. However, the market value of the properties was not necessarily relevant as even if the property market falls, the amount of (and quality of) rental income receivable should not be affected. The property company, Grantchester Holdings PLC ("Grantchester"), has been instrumental in changing attitudes to value and removing the over-reliance placed on loan to value covenants. In fact in the GBP 100 million debenture stock issued in September 1997 by Grantchester there was no loan to value covenant at all.
The second ratio compares the future income receivable (i.e. the rent) with the cost of servicing financing over given periods. With a floating rate loan it is easier to test this on a historic basis, i.e. look back on, say, a quarterly basis to compare the rent received during the previous quarter with the actual interest cost incurred during that period than it is to calculate this ratio on a forward-looking/projected basis. Sometimes that ratio is tested on a forward looking basis, i.e. looking at the next quarter or perhaps next year's rent and financing costs for the relevant period. However, unless the interest rate is fixed, for instance if the financing is raised through a fixed rate bond or the interest rate exposure is effectively fixed through a separate hedging agreement, then this will involve an assumption in relation to future interest rates, most likely that they will remain constant. Even in relatively short-term commercial bank financing, the fixing of interest rates through separate hedging agreements is the norm. Another measure of value?
How then did Grantchester persuade the investors that they were protected in the absence of a loan to value covenant? A concept, mainly seen in the field of project finance, was adapted to achieve the result.
In project finance, the lenders focus on the value of the future receivables, in other words the cashflow the project is projected to provide, rather the value of the assets per-se when they structure a transaction. Similar analysis apply for securitisation techniques, e.g. Highbury Finance & Dragon Finance transactions for J. Sainsbury plc. This is brought into focus when one considers that the assets are likely to be handed back to the state sector at the end of the concession period and so have no intrinsic "value" whatsoever, the only value being the income that they generate during the life of the concession. The "loan life cover ratio" in project finance compares the net present value of the future receivables to the total debt owing. The net present value is calculated using a discounting mechanism to take account of the timing of receipt reflecting the fact that, for instance, GBP 10 received in a year's time is worth less than GBP 10 received today.
This is clearly a far more sophisticated approach than the bank simply looking at a valuation and using a calculator to calculate, say, 60 percent of open market value and deciding that that is how much it feels comfortable in lending. However, in project finance transactions distributions to shareholders are tightly controlled and only permitted when, amongst other tests, financial ratios pass agreed lock-up levels. Distributions in this context include any return to the investors, whether by way of dividend or interest on or repayment of subordinated loans provided by shareholders. Lock-up levels will require a range of reserve accounts (including a debt service reserve) to be fully funded and the loan life cover ratio and debt service cover ratio to reach the required levels.
In property financing, by comparison, property companies generally wish to utilise any excess income each year (excess income being anything in excess of the debt service costs for that year) and either distribute it to their shareholders or re-use the proceeds in other parts of their businesses. The approach taken in the Grantchester debenture stock issue was to incorporate a forecast ratio, tested in limited circumstances, which compared the income receivable over each successive period of 12 months until maturity of the stock with the debt service required for each such period. That forecast ratio was tested when Grantchester wished to substitute different assets (see below) and so it would protect the quality of the income on the substitution, a key area of concern for the investors.
The approach of substituting a forecast debt service cover ratio for a loan to value ratio is reliant upon the income stream amortising the debt during the life of the loan or bond or any remaining income stream on the final repayment date having a net present value of no less than the outstanding debt. If and to the extent that at maturity of the loan there is a bullet repayment the lender carries refinancing risk. If the net present value of the receivables under the leases following the maturity of the debt are not sufficient to discharge the outstandings then the lender is taking the risk (whether he realises it or not) that either (a) if the property is sold it will reach less than the outstanding amount of his debt or (b) if the investor seeks to refinance the property he will struggle to find a lender willing to refinance the debt. So, perhaps to that extent, value, i.e. hope value in terms of re-letting and/or property yields falling and so capital values increasing in order to achieve a sale/refinancing sufficient to discharge the final outstandings, is still called for.
Commercial bank finance and capital market finance
Having considered some of the issues in relation to the importance of the market value of the underlying assets, we will now turn to some of the key advantages and disadvantages of commercial bank finance and capital market debt finance and consider whether the inflexibility of debt capital market finance necessarily outweighs its advantages.
What is the fundamental difference in flexibility?
Finance provided by banks is flexible in nature. Banks will fund loans either through their own resources or on the interbank market. There is scope (depending on the terms agreed with the banks) for moneys to be drawn down, repaid and redrawn which allows the amount of the underlying debt to increase and decrease as the investor's needs dictate. In addition, should the investor wish to deal with the underlying assets, whether by entering into a new lease with a tenant, selling the assets and replacing them with others or the like, then consent can be sought to the changes from either a single lender in a bilateral transaction or the group of lenders in a syndicated deal.
The structure adopted in a capital markets debt financing is markedly different. In a capital markets secured bond or debenture stock issue the documentation includes an offering circular, a subscription agreement (under which a merchant bank sponsoring the issue undertakes to sell the bonds issued and to buy any unsold bonds), a trust deed and security documents. The bonds may be issued in either certificated or uncertificated form but in either event a register will be kept of the bondholders from time to time. The bonds are tradable freely and to enhance marketability may well be listed on a Stock Exchange. The bondholders are likely to be a more disparate group than commercial banks in syndicated financing. They are represented by a trustee who has no financial interest in the debt and who is likely to be permitted to make only technical or administrative decisions without asking the bondholders for consent. In any financing of this nature, in contrast with finance provided by a bank on a bilateral basis or a number of banks on well managed syndicated basis, in practical terms it is more difficult to obtain consent to dealing with the assets. If a decision needs to be made by the bondholders because an objective process has not been established in the trust deed to regulate the issue in question, a meeting of the bondholders will need to be held and a resolution approving such dealing passed. Convening such a meeting invariably takes time and, from the property investor's point of view, there is no certainty that consent will be granted. Therefore, one of the inherent difficulties in financing of this nature is ensuring that there is sufficient flexibility to deal with the assets during the term of the bonds (and thus anticipating what the property investor may wish to do with the assets over the following 20 years or so) without requiring the consent of the bondholders. Property investors need to be able to make decisions quickly and with certainty that their lenders are behind them and therefore the limited scope that the trustee has to sanction dealings with the assets can be a source of great difficulty.
Bank debt is usually provided at a floating rate. Interest rates may rise and, all other things being equal, rent receivable will remain constant so that interest rate increases will lead to pressure on the ability of the property's income to service the interest (although the banks will look for a fairly substantial cushion) and possibly to the transaction ceasing to be viable. This interest rate exposure can be, and generally is, hedged through a separate hedging agreement.
On the other hand, capital market debt finance is almost invariably fixed rate. The rate is fixed by, for instance, reference to the gross redemption yield of the nearest equivalent gilt, i.e. a Treasury Stock with a similar term to maturity, together with a margin in the light of the covenant/security strength of the issuer. As gilt rates are usually lower than bank lending rates, capital market debt should be cheaper, whereas from the lenders' point of view, they are able to obtain a better return than from the nearest equivalent gilt. In addition, the rate will be fixed throughout its term which reduces uncertainty. Of course, this can, in practice, work to the property investor's disadvantage because prevailing floating rates may fall in the future leaving his fixed rate looking expensive and lead to a prepayment of the debt being costly due to the application of the Spens formula (see below).
Bank finance for property investment is on the whole are limited to terms of 5 to 7 years, although 10 years is not unheard of. Capital market debt finance is by contrast long-term as the investors are looking for a better rate of return than they would get on a gilt, with typical terms of, say, 20 or 25 years. Prepaying before the contractual maturity date
A desirable aspect of the flexibility inherent in commercial bank finance is the ability to pay off the loan early and prepay the debt. Subject to paying any costs associated with breaking an interest period for the loan and any agreed prepayment fee, prepayment is relatively painless and inexpensive. However, there may be additional costs associated with unwinding separate swap arrangements which should be borne in mind, although it may be possible for these to be retained on a refinancing.
The application of the Spens formula can make prepaying a capital market debt instrument a costly exercise. Before the bonds can be prepaid the borrower has to pay, together with accrued interest and outstanding principal, an additional amount to the bondholders. That amount is calculated using the Spens formula and, broadly speaking, equates to the net present value of all future fixed rate interest that would have been payable but for the prepayment. The additional payment calculated in this manner can be hugely significant and can be a deterrent to refinancing more efficiently elsewhere. From the bondholders' point of view, they would argue that they have bought an instrument equivalent to a gilt but with a slightly poorer covenant and so a margin on the gilt to reflect that additional risk. The gilt could not be prepaid before its contractual maturity so why should bondholders lose out if the borrower chooses to do so?
Is there any way round this for the property investor? There is usually a restriction on the issuer of the bonds buying back the bonds or, if it is permitted to purchase any bonds, the bonds are immediately cancelled when purchased by the issuer. Reasons cited for this include the logical difficulty in owing a debt to oneself and the undesirability, for the bondholders, of the borrower having voting rights in relation to its own bonds. However, it is well worth arguing on behalf of the property investor that whilst the borrower itself should not be able to buy back bonds other than for immediate cancellation, that restriction should not apply to, for instance, other group companies. That would allow the property investor to acquire bonds at the market price from time to time through another group company, which may well be cheaper than buying them back and prepaying through the Spens formula arrangements.
Flexibility falls into at least two areas. First, the ability to deal with management issues and arrangements in respect of the assets and, secondly, the ability to substitute different assets. Given the limitations on the scope of the trustee's ability to sanction and approve matters, it is important to establish clear and objective criteria and procedures. The property investor needs to be able to manage the property assets on a day-to-day basis without involving the trustee or the bondholders. For example, in the Grantchester debenture stock issue, an initial due diligence exercise was carried out in relation to the retail parks provided as security at the outset in terms of their title and leasing structures. Thereafter, a considerable degree of flexibility was introduced by providing that Grantchester could lease by way of "Qualifying Lease", which provided flexibility to lease in an institutionally acceptable manner with a further test applying to the quality (and diversity) of the income pool. In addition, a regime for accepting surrenders of Qualifying Leases (as surrenders could adversely affect the income) was also incorporated. Further, Grantchester retained the flexibility to redevelop the retail parks.
The approach will vary depending on the way in which the property investor operates its business and sees its business developing. However, the fundamental concern must be to ensure that such matters can be carried out without the delays and cost of approaching the bondholders for their sanction, if at all possible.
Flexibility to substitute different assets is also important. When the bond issue is launched, security will be provided over an initial pool of assets. However, unless the property investor is financing a single asset or a handful of key assets, it may well be the case that the investor might want to sell one or more of the properties forming part of the portfolio and replace them with others during the term of the bonds. There are a number of difficulties to resolve.
The sale of an outgoing property rarely coincides with the purchase of its substitute. In capital market debt financing there is no ability to repay the amounts raised and redraw amounts (as can be achieved with a revolving credit facility provided by banks) when a replacement property has been found.
Every piece of property is unique. Square footage, rent per square foot, number of tenants, tenant covenant and price all vary. However, in this type of cashflow driven financing, the bondholders are really buying into the projected value of a company's business as a whole, rather than a specific portfolio.
If a property investor were to need consent from the bondholders to any such dealing, more likely than not by the time consent had been obtained, the opportunity in question would have gone elsewhere.
Accordingly, a clear, objective regime needs to be established for dealing with property disposals and substitutions. To deal with the difficulty of synchronisation, an interest bearing cash cover account should be established so that the proceeds of sale of an outgoing property are provided as security and those proceeds can be re-invested in other property or other assets at a later date.
As mentioned above, in the Grantchester debenture stock issue, the "value" of each site was not relevant given the financial covenants incorporated in the Trust Deed. The forecast ratio (comparing the income receivable over each period of 12 months until maturity of the stock with debt service required for those periods) was to be tested after a substitution had been effected. However, it was recognised that it would not be appropriate for this ratio to be tested until the series of substitutions in question was complete. It might be the case that there would be a number of outgoing properties which would be replaced with others over a period. Accordingly, it was agreed that the forecast ratio would not be tested until any cash cover had been fully utilised or the period for effecting the substitutions had elapsed. In other circumstances, it may be appropriate to deduct the amount of cash in the cash cover account from the amount of the debt in looking at a true loan to value type ratio. The approach in any such financing is to establish the property investor's concerns and aims and seek to document an agreed procedure to counterbalance the protection of the investors.
As a final point, Grantchester were able to nominate different group companies to provide security and could move the retail parks around within the group. This sort of flexibility is becoming increasingly important as the differential between stamp duty on share transfers and stamp duty on property transfers impacts on the property market. In the Grantchester debenture stock issue, a company could become a charging company relatively easily by giving security over the property in question and a floating charge. The charging company structure means that properties can initially be acquired under a bank facility if acquisition and/or development finance is needed, worked up and then refinanced under the bond structure before being traded on, without needing to transfer the property from the single purpose vehicle initially used to acquire it: in doing so, the effect of stamp duty can be mitigated and in addition the management of the individual sites remains straightforward, with the tenants dealing with the same landlord throughout the process.
The question of what value or price to attribute to a piece of property remains a difficult one. However, it is possible to side-step the question and avoid general market changes causing breaches of financial ratios by looking at the cashflow generated.
Further, property investors should consider raising short term finance through bank facilities and longer term finance through the capital markets. By considering the need for flexibility in dealing, some of the rigidity of capital market debt finance can be mitigated.
For further information, please contact Simon Johnston by e-mail at sim[email protected] or by telephone on +44 (0)20 7367 2008.