by Christopher Shores
At the start of the growth in property investment, many investors were favourably disposed towards leasehold property. More recently, a number of property professionals have declared that, as leaseholds are essentially wasting assets, the return of original capital via amortisation could be compared only with the redemption of a gilt. As a result, the lack of a suitably growth-adjusted end value to a gross redemption yield calculation caused leasehold property to underperform freehold property much more seriously than was allowed for in initial yields then in use by the market.
These pronouncements caused not just a reassessment of yields but wholesale flight from leasehold property. To this day, many institutions will not, as policy, acquire leaseholds of any sort, while most will consider only the longest terms for classes of property where freehold opportunities are strictly limited.
Much leasehold property, particularly of medium term (30 to 80 years), is thus seriously undervalued by investors. Undervaluation springs partly from inadequate and incorrect accounting for the amortisation of the interest. Blame for this state of affairs does not lie with accountants so much as with the property profession itself, which has neither fully considered nor advised on how such amortisation should be sensibly managed. Additionally, there is an over-generous assumption regarding potential future capital expenditure associated with freehold property, and ignorance of the benefit of avoiding such expenditure at the end of a lease.
For many years initial yields adopted for freehold property have been predicated upon the need for occasional injections of additional capital. These are not so much to improve the investment as to maintain its performance in both income and capital growth terms. Indeed, how often rebuilding or major refurbishment may be required is crucial to the yield differential between different classes of property when all other factors, such as covenant and position, are equal.
Traditionally, leasehold property has been valued on a dual rate basis, employing an initial yield represented by the net income after payment of ground rent and allowance for amortisation, coupled with a historic (and now unrealistically low) compound interest rate for replacement of original capital. The figure most commonly in use today is 4% pa, which means that a fairly large deduction is made in arriving at the net income. If this deduction — which is made quarterly in advance — is properly reinvested in some asset providing a secure but reasonable return, it will, in fact, compound to a substantially higher sum at the end of the lease period than the original acquisition cost. Since this deduction is being made from the proceeds of property, then it is only logical that its reinvestment performance should represent a part of the performance of the property itself, as does the net rent.
Equally logically, since it is property-produced, it should be reinvested in some investment medium offering a similar return to property. Currently, however, insurance companies when dealing with performance of regular savings plans in unit trusts normally indicate likely performance assuming 8% and 13% pa returns. Consequently, in this attempt to reassess leasehold performance, reinvestment of amortisation deduction at these two rates has been assumed. On longer leases, the difference in the internal rate of return occasioned by applying differing amortisation rates is not large since the deduction for reinvestment is small. However, the shorter the term of lease acquired, the greater the differential becomes.
It should be remembered that the initial yield even on a very long leasehold, which is applied to net income, is somewhat higher than a freehold. Nevertheless, the amortisation deduction made is a constant factor throughout. It is not related to a percentage of rental growth in a leasehold property and so will result in a higher percentage increase on net rent than will be the case with a freehold, given similar overall rental growth rates. Indeed, the higher the amortisation deduction made, the more this will be the case. The resulting increased yield for the duration of the lease goes far to make up for the lower end value.
Alternatively, if rental growth were to be directly comparable with a freehold, then the percentage of the rent received initially — which represents the sinking fund deduction — should remain constant at each increase in income. Effectively, this would increase the amortisation investment at each review or reletting at a similar rate of growth to that reflected by the overall rent. This would have a dramatic effect on the final replacement capital sum achieved without adverse effect on the performance of the basic property investment, compared with a freehold.
To demonstrate this thesis, a number of models of comparable properties have been constructed on both freehold and leasehold assumptions. Whether the assumptions made therein are reasonable is obviously open to individual assessment, but the important factor is that they are common to each.
(1) Prime High Street shop
We will assume that this property will retain its favoured position and that tenants will readily be found in the event of any vacation by the existing occupier, such tenants being prepared at their own cost to undertake shopfitting works. On this basis, the property is expected to be entirely viable for use in 100 years and no provision for major refurbishment, redevelopment or void is necessary. A growth rate applicable to the rental value has been employed at 10% pa; a general inflation rate of 6% pa.
It has also been assumed that the property is rack rented at the date of acquisition. The leases under which it is held are assumed to be fully repairing and insuring with upwards-only rent reviews at five-yearly intervals.
For the freehold calculation, an initial capitalisation rate of 4.5% has been used to assess the initial cost, and it has been assumed that this will still be an acceptable return in calculating the end value.
This shows an internal rate of return of 14.0628% unadjusted for inflation but, more relevantly, if allowance is made for inflation at an average annual rate of 6%, the adjusted IRR is 7.6%. The latter basis has been adopted for the further examples which follow.
If we now assume that the interest to be acquired is a headleasehold, with 50 years unexpired, and at a ground rent geared to 10% of rack rental value, an appropriate capitalising yield at the present time is likely to be of the order of 8% and 4% (dual rate).
This example demonstrates a remarkable adjusted IRR of 10.83% — even at the base 4% amortisation assumption. At the more appropriate 8% and 13% rates, the IRRs increase to 11.1% and 11.175% respectively. Clearly there would seem to be scope here for a reduction in the capitalising yield, while still representing a most acceptable alternative investment to the freehold.
(2) Prime city office property
Freehold and rack rented as before, but assuming a refurbishment at years 25 and 75, and redevelopment at years 50 and 100, in order to maintain rental growth at a level of 10% pa. Initial yield is assumed to be 5.5%. As a comparison, this calculation sees refurbishment as being necessary only at years 50 and 100, but with rental growth thereby being limited to 8% pa, other factors remaining unchanged.
If growth is maintained by the level of further expenditure indicated, the adjusted IRR is 8.26%. However, if a lower overall rate of growth is accepted for a lesser level of future additional input, the adjusted IRR falls to 6.7%.
This same property is then considered on the basis that it is held on a 99-year lease with a ground rent which is reviewable co-terminously with the occupational leases, and is geared to 20% of the rental values. Because it is leasehold, it will not be necessary to provide either for refurbishment or redevelopment at the end of the term, although refurbishment at years 25 and 75 and redevelopment at year 50 has been built in. With all other assumptions unchanged, an initial yield based on dual rate tables at 6.5% and 4% sinking fund is adopted, with no provision for tax on the income from reinvested amortisation.
If amortised at the “face” 4%, the adjusted IRR is 7.16%. However, if the return on the reinvested sinking fund is taken at 8%, the adjusted IRR increases to 7.3% and at 13%, to 7.46%. It will be noted that these figures are approximately midway between the two freehold scenarios adopted. Since the yields on amortisation employed are themselves not inflation adjusted, it would seem reasonable to assume that the IRR demonstrated on these assumptions will be at the lower end of that achievable.
The examples would seem to indicate that properly managed and accounted-for leasehold properties will perform at least as well as, and frequently better than, a similar freehold. Particularly for medium-term leases there would appear good reason for reducing the margin of capital differential between the two.
There is one important rider to this advocacy of the underrated leasehold. In considering leasehold property, careful attention does need to be given to the terms of the headlease. While these investments may be considered frequently to be more attractive than current doctrine may hold, they can have flaws and dangers which, if understood, may readily be avoided.
Because an investor may wish to carry out works of refurbishment or redevelopment which require the property to be vacant for a time, the head rent should never represent too high a proportion of the total rental value of the property, since the continuing requirement to pay this rent even when no occupational income is forthcoming can increase the costs of redevelopment and, in the case of a protracted void, can place a burden on the investor which may well destroy the performance of the investment over a long period.
Ideally, the investor should guarantee a basic return only during periods of vacancy, and any share of rental growth paid to the freeholder in excess of this should be linked to actual rental income. For this reason, it is more desirable that all such equity-sharing rents be expressed in terms of rents received, rather than rents receivable.
As a guideline, basic ground rents should never exceed 20% of rack rent value, and ideally should be 12% or lower, particularly in the case of industrial property where continuity of income owing to frequent reconstruction is the least certain. Provision should always be incorporated for the headlessee to be able to undertake redevelopment or refurbishment as reasonably necessary throughout the term without undue interference or control by the freeholder.
These comments have related specifically to leaseholds held on modern terms. Many older long leaseholds still exist which are held on terms now very favourable to the headlessee. Typically, these may include a very low ground rent which has no provision for review, or where reviews are only very low percentages at infrequent periods. Such interests can often be acquired very favourably and offer scope either to purchase the freehold or to restructure the lease on to a modern basis, but at a cost to the headlessee which further enhances the performance of the asset when compared with a direct freehold acquisition. Such possibilities are well worthy of consideration by any institutional investor seriously considering property investment on anything other than the most “hands-off” basis.
