The research register for this journal is available at The current issue and full text archive of this

f this journal is available at

http://www.mcbup.com/research_registers/jpif.asp http://www.emerald-library.com

ACADEMIC PAPERS Academic papers:

Uncertainty in
Uncertainty in property property
valuation
valuation
13
The nature and relevance of uncertainty
and how it might be measured and Received October 1998
Revised December 1999
reported
Michael Mallinson and Nick French
Department of Land Management and Development, The University of
Reading, Reading, Berkshire, UK
Keywords Uncertainty, Market value, Valuation
Abstract There will always be a degree of uncertainty in any valuation, but it should be
incumbent upon the valuer to report ``abnormal uncertainty''. This arises when some particular
condition of the market or the property leads to the valuer being unable to value with the
confidence of accuracy which might normally be expected. Abnormal uncertainty now features in
the RICS Appraisal and Valuation Manual and later in this paper we consider how the valuer
might identify and measure the degree of abnormal uncertainty. But this paper is predominantly
concerned with ``normal uncertainty'', which hereafter we will term only as ``uncertainty''. The
thesis of this paper is that uncertainty is a real and universal phenomenon in valuation. The
sources of uncertainty are rational and can be identified. They can be described in a practical
manner, and, above all, the process of identification and description will greatly assist many
clients, and will improve the content and the credibility of the valuer's work. Common professional
standards and methods should be developed for measuring and expressing valuation uncertainty
(Recommendation 34, Mallinson Report, RICS, 1994).

Introduction
In March 1994 the Mallinson Working Party on commercial property
valuations produced its report outlining a number of initiatives which the RICS
should undertake to help improve the standing of the valuation surveyor
within the business world.
A number of these recommendations concerned improving the
communication process between the client and the valuer and these
recommendations have formed the foundation of the new provisions contained
within the revised Red Book (RICS Appraisal and Valuation Manual, 1996).
However, the Red Book is principally concerned with the purpose of the
valuation; the basis of the valuation and the form of reporting the valuation to
the client. It does not fully address the question of quantitative method or
indeed the limitations of the valuation figure.
The Mallinson Report argued that all valuations are uncertain. A valuation Journal of Property Investment &
figure is an individual valuer's estimate of the exchange price in the Finance, Vol. 18 No. 1, 2000,
pp. 13-32. MCB University Press,
marketplace. They are not a statement of discerned, objective fact, but rather 1463-578X
JPIF deduced assessments that give an expert's opinion. Despite this, clients and
18,1 third parties who intend to use a valuation as a basis for action are inclined to
treat it as if it were fact.
One of the principal tenets of The Mallinson Report was that valuers should,
when necessary, expand on not just the details about the property, but also on
the valuation itself; its dynamics; its relativities and its uncertainties.
14 There will always be a degree of uncertainty in any valuation, but it should
be incumbent upon the valuer to report ``abnormal uncertainty''. This arises
when some particular condition of the market or the property leads to the
valuer being unable to value with the confidence of accuracy which might
normally be expected. Abnormal uncertainty now features in the RICS
Appraisal and Valuation Manual and later in this paper we consider how the
valuer might identify and measure the degree of abnormal uncertainty. But this
paper is predominantly concerned with ``normal uncertainty'', which hereafter
we will term only as ``uncertainty''.
The thesis of this paper is that uncertainty is a real and universal
phenomenon in valuation. The sources of uncertainty are rational and can be
identified. They can be described in a practical manner, and, above all, the
process of identification and description will greatly assist many clients, and
will improve the content and the credibility of the valuer's work.

What is ``uncertainty''?
When preparing a valuation, the valuer is seeking to estimate market price; the
price (in either capital or rental terms) at which the property might be expected
to transact on the basis of the given assumptions. In reaching his/her
judgement, uncertainty will arise in the valuer's mind, owing to the difficulty
either of assessing the market itself or of assessing how the market would price
the particularities of the subject property.
Under the bases of valuation set out in the RICS Appraisal and Valuation
Manual (e.g. market value (MV) or open market value (OMV)), the valuer is
instructed to view the transaction through the eyes of a hypothetical buyer.
The valuer must consider all possible buyers in the market in order to identify
the ``best'' price likely to be forthcoming. In performing this task the valuer will
be uncertain about the current availability of buyers, their current attitude to
price, and what they would make of the particular property.
Against this background, in even the simplest of valuations there are likely
to be a number of variables that the valuer must assess. For example, in a
vacant possession house valuation, even if the house is similar to others which
have been sold recently, the valuer must assess, through the eyes of the
hypothetical purchaser, slight differences in location, the time since the last
transaction, differences in standards of fitting, etc. For more complex
properties the number of variables will multiply. In order to produce the
valuation, the valuer must weigh all the variables, using his/her skill and
experience, and decide on the most probable conclusion.
 The valuer may or may not prove to have been ``accurate'' if a transaction Academic papers:
actually occurs or if an observer attempts to measure accuracy. There are great Uncertainty in
difficulties in identifying accuracy in property valuation. From the perspective property
of the courts, judges commonly refer to the ``correct figure''. However, this is
misleading. They actually are referring to a figure somewhere within a range
valuation
that a number of competent valuers would have reached. They will have no
idea whether such a figure would, if tested by an actual sale at that point in 15
time, have proved ``correct'' or ``accurate''.
For the purposes of this paper we are seeking to identify the substance and
the characteristics of the uncertainty which lies in the valuer's mind as he/she
attempts to assess the hypothetical purchaser's view of the variables involved.
This is of interest because the valuer is expert in the field and we are talking of
his/her view of the variables which are driving, or will drive, price. In many
circumstances this would be an insight of great value to clients.
In this paper, therefore, we are trying to identify uncertainty in the valuer's
mind when he/she is attempting to determine an estimate of market price.

What is the relevance of uncertainty?

If uncertainty in the mind of the valuer is a fact of valuation, then its open
acknowledgement and measurement will only be relevant to the process if it
either improves the position of the client or improves the credibility of the
valuer. Even the most naõÈve client or observer is unlikely, at present, to think
that a valuation is accurate to the nearest pound. The word ``valuation'' already,
therefore, carries with it an accepted aura of uncertainty. The sceptic may thus
query whether further identification will bring relevant benefits.
Despite the current acceptance, in principle, of the uncertainty of a valuation,
a long valuation report with a single figure at the end of it is potentially
misleading and unhelpful to the client. It also puts the valuer unnecessarily on
the defensive if the property transacts at a different figure, or the valuation is
challenged in some way. Generally valuers are wary of including details of
their thought processes within the report itself as they feel it leaves them open
to criticism in the event of the valuation being challenged. However, we would
argue that an open revelation of uncertainty would contain information
potentially valuable to both client and valuers. Openness about uncertainty, in
a professionally organised manner, will be good for both client and valuer.
At first, clients may see any element of uncertainty as being entirely against
their interest. They wish valuers to produce their figures with complete
accuracy and no uncertainty. However, once it is accepted that uncertainty
exists, then a wise client may prefer to hear about it, provided that it is reported
in an organised and helpful way. A lender who is considering whether or not to
exercise the power of sale against a defaulting borrower may seek a valuation
before deciding whether to embark upon this course. Any variability in the
valuation figure may be of considerable importance in the decision. Equally, an
investor or owner-occupier considering buying or selling may seek a valuation
JPIF before embarking on the effort or expense of the transaction. Again, an
18,1 understanding of the valuer's uncertainties may give a key insight into the
desirability of proceeding.
It may be argued that, in some cases, the client has no interest whatever in
uncertainty. This view may be given when the client is not intending any
action, but purely requires a figure, as a benchmark or a neutral observation,
16 for the purposes of accounts perhaps, or the calculation of a unit price. This
view, commonly expressed, is too superficial. The client may indeed be only
interested in the figure because no action is intended, but others may intend to
act upon the basis of the valuation, to buy or sell shares or units perhaps. They
would indeed be interested in the valuer's uncertainties.
In some cases it must be accepted that the importation and the expression of
uncertainty might seem to raise more problems than it solves. A valuation for
probate, which included uncertainty, might give an unwelcome opportunity for
the Inland Revenue to challenge. Similarly, with a valuation given in court or as
part of a compensation case, or as an independent, expert adjudication between
two parties, the possibility of challenge may be increased if uncertainty were
expressed. This problem may be less than it seems. In the first place, both sides
to an argument would have to express their uncertainties if a challenge were to
carry weight. But, and more importantly, any difficulty would be more than
outweighed if the single figure given as ``the valuation'' is seen to carry the
authority of the valuer and that authority is enhanced by the expression of
uncertainty. If the valuer is seen to have considered a number of possibilities, to
have used his/her skill to choose between them, but accepts fallibility within
limits then the valuer is credible, and the client will benefit from that
credibility.
There is a difference between the uncertainty within the process of valuation
by a particular valuer and the variance of value between valuers. The courts
are concerned, not with the individual's uncertainty, but how his or her figure
compares with figures produced by other valuers, how far from a consensus
figure a valuer can stray without being negligent. Variance and acceptable
tolerance are, in themselves, variable. They will vary from property to
property, from situation to situation, and from market to market. If the
profession professed uncertainty and a knowledge of it, then the correct
foundation for that profession must be the uncertainty of each valuation.
Understanding and identifying uncertainty is potentially highly relevant to
clients and to valuers.

How does uncertainty manifest itself?

Figure 1 is an illustration of a ``normal' distribution. The horizontal axis
represents the quantum of value and the vertical axis the probability of any
individual figure, with the total line amounting to 100 per cent. Thus the curve
represents a 100 per cent probability of the value lying between Va and Vz,
 Academic papers:
Uncertainty in
property
valuation

17

Figure 1.
Illustration of a ``normal''
distribution

with the highest probability of an individual figure of V, which has a

probability of Pz per cent. As drawn, the curve also shows that the probability
of the value lying above or below V is equally distributed either side of V.
Let us start our consideration of uncertainty away from property. If a
stockbroker is asked to ``value'' a line of quoted shares at a certain date for
account purposes, this is the equivalent of market value (MV) in property
terms. The underlying ``assumptions'' made by the stockbroker will be, in
effect, the same as for MV in a property valuation. However, the stock
valuation is rather simpler. All the stockbroker needs to do is to refer to the
Stock Exchange list on the desired day and value the line of stock at the
mid-market price given. The stock valuation has no uncertainty. The ``curve''
will be a single point with 100 per cent probability. There is no ``information'' in
the stock valuation other than the figure itself.
If the line of stock is larger than a market maker will take then the
uncertainty will grow. The broker must now assess a number of variables, the
discount/premium market makers might seek for a larger line, the greater
chances of market movement if the line has to be fed into the market over time,
and so on. The ``curve'' will start to carry a meaningful spread, and thus, even in
a highly liquid market, a bell curve can start to emerge.
Referring again to Figure 1, it can be seen that there is an equal distribution
either side of a most probable figure. In most markets this will not be the case.
Markets are constantly on the move. The valuer, to the extent that he/she has
uncertainty, will feel it more to one side of the most probable figure than the
other. If the market is rising it may be that most probabilities will be skewed on
the ``up'' side, as in Figure 2. Remember that we are not identifying the
probability of accuracy but the valuer's personal uncertainties in the light of his
JPIF
18,1

18

Figure 2.

or her knowledge of the market. However, such a distribution need not only
reflect the dynamics of the market. It may also reflect the caution or bullishness
of the valuer.
Thus, Figure 2 might suggest that the market is rising, but it might also
reflect that the valuer has chosen to be cautious. That caution might derive
from a sense that the market may be turning or because of some particular
uncertainty deriving from some aspect of the property itself. A curve with the
opposite skewness, as in Figure 3, may suggest a degree of bullishness about
the figure, perhaps deriving from some knowledge of a real potential purchaser,
or it might spring from an observed falling market. As will be seen, the shape of
the curve may contain considerable ``information''.

Figure 3.
The curves contain information not only about the market, but also about Academic papers:
where the valuer chooses to place the figure of valuation within that market. Uncertainty in
Note that, in both curves, it may well be that Va and Vz will carry the same property
figure. The shape of the curve derives from the placing of V. To be clear, two
valuers, valuing the same asset, reaching the same conclusions on the
valuation
variables, and reading and observing the same market, may still choose a
different figure of value. If they do so, the shape of the curve around their 19
valuation will be different. The curves thus contain information about the
valuation within the market that may be of considerable help to a client.
A non-property example would be the sale of whole companies. There is no
``market'' for whole companies. All we do know is that they commonly change
hands at prices materially different from the sum of the marginal prices. In
deciding to buy the company, a potential purchaser will have an idea of its
``worth'' to them, and they will seek valuation advice on the price they might
have to pay in the market. The valuer will perform a calculation which will
contain considerable uncertainty; uncertainty which can be illustrated by the
bell curve. Before embarking on the potentially expensive, and certainly
revealing, exercise of attempting a purchase it will be crucial to understand
where the valuation advice you have received stands in relation to its
uncertainties. If the company is ``worth'' more to the purchaser than Vz, then the
only concern, but still a real concern, will be to buy it as cheaply as possible.
But if ``worth'' lies somewhere between Va and Vz, understanding the valuer's
uncertainties will be crucial in deciding the wisdom of launching a bid. If the
target company is privately owned, the potential purchaser will be free to
approach the owners in whatever way seems best, and to make a bid at any
level, striking the price where both parties agree. The distortion of the curve in
the eyes of the advising valuer will reflect the likely time for reaching a deal,
taking account of the general condition of the market, and the negotiating
strengths and weaknesses of each party.
If the company is quoted, however, things become very different. Takeover
Panel rules apply. This has two effects. The curve may become more extended
because of the risk that the bid will flush out other possible buyers. Second, the
bid (B in Figure 4) sets the price. For all practical purposes that becomes a
minimum price. The choice of that figure will cut off all probabilities between
Va and B, leaving only the probabilities above B. An understanding of the
curve which the valuation advisers have in their minds may prove crucial to
the purchaser in either securing the deal, or getting it at the most advantageous
price.
The crucial point is understanding the views that lie behind your adviser's
figure, views that will be revealed by an organised description of uncertainty.
A figure of V, however well calculated, however authoritative, however well
protected by professional indemnity policies, brings very little information that
is really valuable. The manifestation of uncertainty will depend on, and its
description will reveal, causation. Causation will expose valuable information.
JPIF
18,1

20

Figure 4.

Uncertainty in property valuation

In the previous section we have demonstrated that ``normal uncertainty'' is not
unique to property. It will occur in all valuations that are not a simple recording
of referable fact such as the production of MV for a line of quoted shares. The
issues and principles discussed apply equally to property, but there are some
additional issues.
Even the valuation of a property in perfect condition in a centrally prime
location, let at a provable open market rental value (OMRV) will generate a
considerable spread of views (Va-V-Vz, reference ). The bell curve distribution
is intrinsic in the thought process behind every property valuation. The
underlying process of valuation is the same for property as it is for any other
asset.
However, property does have certain peculiarities that occur less frequently
in other markets. One general factor the property valuer needs to consider
which will not arise, or not arise materially, in any other market, is the
likelihood of the existence of a buyer at the time of the valuation. In principle,
particularly for quoted shares but probably for all assets traded at the margin,
there are a large number of buyers always potentially available. Indeed their
absence would be a major signal for ``abnormal uncertainty''. The quantity and
enthusiasm of the purchasers is only a product of price, and their views of this
are on display in market movements minute by minute. This situation arises
because of the marginal nature of the market, because of its liquidity and
transparency, the existence of derivatives, etc. All these features encourage
arbitraging and the taking of long and short positions. None of this applies to
property.
It is not uncommon that great difficulty can arise in finding a buyer for a
particular property at a particular time; witness the frequency with which no
real bids are received at an auction or the problems house owners sometimes
experience in finding a buyer. This situation may arise for many reasons; Academic papers:
perhaps potential buyers are currently digesting, either in financial terms or Uncertainty in
personal attention, other transactions. Whatever the cause, if truly it can be property
seen as temporary, it may not impinge too much on MV. Therefore, in both valuation
setting the valuation and pondering the bell curve, a property valuer will have
to take into account, insofar as he/she is able to do so, the current situation of
21
actual buyers. It is appreciated that the valuation is concerned with a
hypothetical buyer but that the hypothetical buyer must not be mythical.
This issue can be illustrated with a real example. The working party that
prepared The Mallinson Report held a number of interviews with clients and
valuers. We put before them the result of an actual open market tender
(Table I) for a small group of normal, clean investment properties.
There were no obvious ``angles'' and none of the tenderers was a special
purchaser. The interviewees were asked to name the figure for which the valuer
should have been striving in an attempt to identify MV. Interestingly, the
valuers, relying on the adjective ``best'' in the definition, went for the top of the
range, whereas most clients, except the property companies, sought comfort,
and ``repeatability'' around £1,300,000 to £1,400,000. This array of bids does
not equate to Va-V-Vz. They are actual bids and it is unlikely that a valuer
would have thought it necessary to consider quite such a wide range. But,
assuming that the valuer's thought processes did indeed reflect in some way
the attitudes of these potential buyers, the position of V and the shape of the
curve which resulted from that decision would have influenced the interest of
the top bidders.
Further, it will be quite common for the valuer to have in mind potential
actual bidders when setting the figure. This example demonstrates not only the
potential value to clients in understanding the valuer's uncertainty in a
structured manner, which we may assume will reflect to some degree the

Top ± highest bid £1,950,000

Second £1,700,000
Third £1,550,000
Fourth £1,410,000
Fifth £1,380,380
Sixth £1,305,000
Seventh £1,300,000
Eighth £1,275,000
Ninth £1,260,110
Tenth £1,260,000
11th £1,205,247
12th £1,200,000
13th £1,120,000
14th £1,095,000 Table I.
Remainder + 4 below Bids at tender
JPIF reality shown by the bids, but also the importance of the valuer understanding
18,1 actual buyer availability. The problems of illiquidity and of observing the
depth of the market are additional issues for the property valuer.

Influences on the property curve

The central tenet of this paper is that the identification of uncertainty manifest
22 in the bell curve contains information which is of value to clients, and which is
relevant if valuations are to be properly understood.
The shape of the curve in any particular case will be influenced by many
factors, and it is perhaps worth identifying a few of the more general ones in
order to demonstrate not only that the curve will reflect important concerns,
but also that interpretation of the curve will help the client to appreciate more
fully the advice being given. While it would be untrue to say that such advice is
not forthcoming at present, it is not commonly given and when it is it lacks the
credibility which would come from a common professional approach.
For illustration seven influences will be discussed, but this list is indicative
and not exhaustive.

Liquidity
The liquidity of an asset will derive not only from the nature of the market in
which it is traded, but also from the characteristics of the asset within that
market. Property has very variable liquidity, from prime investments at the
most liquid end of the spectrum to specialised properties that are rarely traded.
The greater the liquidity the less the uncertainty in the mind of the valuer.
Comparables are likely to be more common and closer in time to the valuation;
liquidity will therefore influence both Va-Vz and Pa-Pz; the more liquid the
narrower the range of the former, and the higher the value given to Pz. If any
asset, as a type, or because of some characteristic of the property, is rarely
traded at the time of the valuation, the valuer will be less certain of the total
range within which a transaction might occur and of the likelihood of one
figure rather than another. The curve will therefore tend to be flatter with
perhaps less skew.

Depth of the market

As illustrated previously, concerns over depth will influence not only the
flatness of the curve as the probability of any one figure will be low and the
potential range may be greater, but also its skewness. Doubts about the
availability of high payers will tend to weight probabilities below the
valuation.

Availability of evidence
This can be differentiated from liquidity, especially for property due to its
heterogeneous nature. A key element of any valuation to market price is the
availability of evidence from transactions both near in time and near in type,
nature and location. While it is appreciated that comparables are only part of
the process and that experience and accurate observation of the asset itself are Academic papers:
also important, it must still be true that the absence or remoteness of evidence Uncertainty in
will increase the valuer's uncertainty. Dependent on the nature of the paucity property
this might influence Va-Vz, but it is more likely to influence the value of P valuation
around the most probable area, flattening the top of the curve only.

Intrinsic uncertainties of the asset itself 23

Given the legal, physical and geographic uniqueness of each property,
individual features may give rise to increased uncertainty, uncertainty about
quite how the market would interpret and price an identified feature. It may
well be that particular features are given more weight in some conditions of the
market and less in others. For example, a minor defect in a rent review clause
may be almost overlooked when buyers are keen, but assume greater
importance when the market is weak. Such uncertainties, if only of marginal
concern, will tend to reduce the value of P around the most probable area. If
they are more fundamental they may extend the range of Va-Vz, and indeed
may distort to the extent that ``abnormality'' emerges.

Dynamics of the market

The dynamics of the market will generally be reflected in the skewness of the
curve, but in a comparative rather than absolute way. In a rapidly rising
market it is, prima facie, more likely that uncertainty will be weighted on the
upside and in a weak one on the downside.

Assumptions
These will be important to the curve. Sometimes a special assumption may
reduce uncertainty, if a valuer is instructed to ``assume away'' a difficulty for
example.

The valuer's choice

The greatest influence of all on the curve will be the final choice of the single
figure valuation. This is as it should be. We are seeking to capture the valuer's
uncertainty and, while he/she will have generalised uncertainty as the options
are considered, it will not really crystallise until the choice of value is made.
The valuer's choice will influence only skewness. If, in pursuing ``best'' price,
he/she tends to go near to the top of the full range of possible values, it is likely
that, while the highest level of P will lie at the chosen valuation, there will be
more possibilities below than above. The use of the uncertainty curve will give
the valuer the chance to put the valuation choice into an explicit context.

Establishing a property curve

The bell curve for any particular property valuation will be the product of two
variables, value and probability. It would be perfectly possible to generate a
JPIF model that took account of more variables than those discussed in the previous
18,1 section. However, to illustrate this particular point, we will deal with only two
variables, which represent the two axes of the graph.

Setting values
Figures of potential value to be included in the curve will be the product of the
24 valuer's normal calculations, implicit all risks yield, explicit DCF, profit based,
direct comparison, etc. However, if a curve is to be produced valuers will have
to change their approach somewhat. With present practice it will be usual for
the valuer to view all the variables in whatever calculation it is intended to
perform, and set them at the figures which seem most apt. The calculation will
then be performed using those figures. It will be normal also for the valuer to
carry out a few ``what if?'' calculations to establish sensitivity, and the final
judgement of value will be made between a number, but a limited number, of
calculations.
Two points are worth noting. First, each calculation, although producing a
single figure answer, will not be viewed by the valuer as being precise; it will
carry its own aura of uncertainty and the valuer will use judgement to round or
adjust each figure. Even if the market were precisely echoing the variables used
by the valuer, rounding would produce variability. The valuer will be
conscious of this and attempt to allow for it, but in essence each calculation is
no more than an ``observation'' of part of a probable range of answers. The
valuer will attempt to master this uncertainty in the eventual choice of figure,
but proper mastery is impossible, and compromise inevitable. Second, when
reviewing the outcome of the range of observations made, the valuer will, at
present, be focused on the definition of value, the ``best price''. It is his/her task
to find that figure and in struggling to do so the less probable figures will be
discarded. This is inherent in the present methodology of producing valuations.
However, these figures, although less probable in the view of the valuer, are
perfectly possible. But at present they are discarded and feature nowhere in the
eventual valuation.
Yet, it is likely that the sum of the probabilities of all these alternative
answers will be greater than the probability of the single figure produced by
the valuer. This is likely to be true even if one makes the ``observations'' into
ranges. This is a substantial lacuna in the advice that a valuer conventionally
gives and leads to misunderstanding.
Having identified the range of the variables it is possible to produce a matrix
of possible single figure valuations. Provided that the extremes have been set
correctly and realistically, all the possibilities will have been captured. The
``correct'' value is certain, or virtually certain, to lie somewhere within the
matrix. Unlike with the conventional approach, these possibilities will not be
discarded, but all will be considered and given a probability.
Setting the probabilities Academic papers:
To attempt to set a probability for each single figure would be a considerable Uncertainty in
task, nor would it be particularly fruitful as it would not necessarily reflect the property
thought processes of the valuer. It would be better to isolate a series of ranges,
valuation
or ``observations'' each in the form of a range. These observations might relate
directly to the results from using one set of variables, or they might be chosen
more subjectively, exercising judgement. Each of these observations then has a 25
probability ascribed to it. Each observation is possible, but some will be more
probable than others. These probabilities will be strongly influenced by the
view of the ``influences'' discussed in the previous section and by the valuer's
view of the particular features of the property.
A valuer should have little difficulty in reaching an opinion on the order of
probability of each observation, cascading away from the ``most probable'' in
the form of the bell curve. However, the curve will no longer be a curve. Because
each observation is a range and not a single figure a true curve is not plottable,
but rather a ``fuzzy curve'' or ladder emerges as in Figure 5.
The process of reaching the observations which, in total, amount to Va-V-
Vz is not particularly difficult for the valuer and is little more than an extension
of present practice. But the setting of numbers to the probabilities Pa-Pz would
be wholly new. This is illustrated in Table II.
In Table II the probabilities are entirely subjective. They are what the valuer
thinks. In many statistical exercises in which probabilities are used they are set
in a more quantitative or econometric manner. They are calculated in a way
which gives them some degree of mathematical integrity. We do not believe
that the majority of valuers can achieve this, nor is that the objective. The
purpose of the exercise is to identify the valuer's uncertainty which is, by
definition, entirely subjective.

Figure 5.
JPIF Assumptions
18,1 1. Retail warehouse of 30,000 sq. ft net.
2. Let to good convenant for 25 years from 1988 on FRI terms with upward only reviews
every five years. Current rent, set 1993, £210,000.
3. Structure clean. Adequate car parking. Adequate prominence.
4. Small amount of methane contamination on one part of car park. Might lead to a price
26 reduction of 0 per cent to 5 per cent.
5. Repairing covenant badly worded, but appears to have caused no problems in 1993.
Might lead to price reduction of 0 per cent to 5 per cent.
6. The investment market is firm for this type of asset with a number of known buyers.
They appear to be seeking a target return (equated yield) of between 8 per cent and 9 per
cent.
7. The rental market, having been strong, has levelled off in the last 12 months. Tenants
appear willing to accept between £7.50 and £8.50 at review, with examples above for
new lettings and a few arbitrations below.
Matrix valuation: £000s
OMRV per square foot
Equated yield 7.25 7.50 7.75 8.00 8.25 8.50 8.75
8 per cent 2,511 2,689 2,757 2,845 2,923 3,001 3,079
8.25 per cent 2,531 2,606 2,681 2,756 2,831 2,906 2,981
8.5 per cent 2.464 2,536 2,609 2,682 2,754 2,827 2,899
8.75 per cent 2,392 2,452 2,533 2,603 2,673 2,743 2,813
9 per cent 2,325 2,393 2,461 2,529 2,597 2,664 2,732
Notes: contamination reduction ± up to between 116 and 150; lease reduction ± up to
between 116 and 150.
Conclusions
1. The proper provable OMRV is, in the view of the valuer, £8 pfs and the market would
look for an EY of 8 per cent. Neither of the defects is serious, but it would be expected
that a price reduction of around £100,000 would need to be agreed to get the contract
signed. Therefore settle MV £2,575,000, i.e. around the centre of the matrix, adjusted for
the reduction.
2. Because the market is pretty liquid for this type of asset, the valuer is confident of the
figure. Therefore adopt a narrow most likely range from £2,500,000 to £2,650,000 and
give this a probability of 40 per cent.
3. Because the market is quite firm it is more likely that a price would exceed £2,650,000
than fall short of £2,500,000. Thus weight probabilities accordingly. But the defects are
there, and if the keener buyers choose to pass this one a price towards £2,300,000 is not
impossible, but below that is unlikely with only a small chance. We have some chance of
an optimist and a figure of up to £2,900,000 is coherent on a number of assumptions, but
more than that is considered to be unachievable.
4. Viewing the matrix, a good number of answers fall between £2,400,000 and £2,750,000.
The probabilities will therefore be concentrated within this band, which is 6.8 per cent
either side of the median. This tends towards a normal distribution.

Table II. Probabilities

A worked example to 2,025,000-2,300,000 2 per cent
identify and record 2,300,000-2,400,000 5 per cent
uncertainty (continued)
2,400,000-2,500,000 15 per cent Academic papers:
2,500,000-2,650,000 40 per cent Uncertainty in
2,650,000-2,750,000 20 per cent property
2,750,000-2,850,000 10 per cent
2,850,000-2,950,000 5 per cent
valuation
2,950,000-3,079,000 3 per cent
Report to client 27
MV £2,575,000
Probabilities 2,500,000-2,650,000 40 per cent
2,400,000-2,750,000 75 per cent
2,025,000-3,079,000 100 per cent
Skew less than 2,500,000 22 per cent
more than 2,650,000 33 per cent
This dispersal of probabilities is quite normal and you should feel able to rely on the MV
within normal margins. You will note that it is more likely that the MV figure would be
exceeded rather than undershot. Table II.

If we accept subjectivity, are there any rules by which it might be sensibly

constrained? There seem to be three options for any professional rules.
First, it could be decided, as a matter of professional practice, that each
observation should have the same aura or range (e.g. 2.5 per cent either side
of a central figure). The probability attached to each observation would then
be allowed to vary at the valuer's discretion. Second, the rule could be
reversed. That is to say each observation would be required to be in
constant steps of percentage probability (e.g. 5 per cent steps, or perhaps in
steps of 10 per cent, 7.5 per cent, 5 per cent, 2.5 per cent, etc. in a flattening
curve), but the range of each observation would vary at the valuer's
discretion. Both these alternatives would give a structure and discipline to
the process. The third option is to allow the valuer complete discretion in
setting both the range of each observation and the probability steps. This
latter option fits more closely with current valuation techniques, it gives
valuers freedom in colouring their professional view, and finally it is more
appropriate to the great diversity of situations which arise in property
valuation. With continued use and some degree of shared experience,
subjectivity will be contained to the extent that the description of
uncertainty will acquire some degree of common shape, matched to the
demands made upon valuers.

How should uncertainty be described?

The ladder in Figure 5 contains comprehensive information on the valuer's
view of value within a subjective framework, limited only by the definition of
value being considered with its attendant assumptions. To a person intending
to use the valuation as one component in a decision an understanding of the
``comprehensive information'' will be of potential importance. How can this
JPIF information be described and conveyed? There are two difficulties to overcome,
18,1 the potential complexity of the information and the common simplicity of the
client. This, of course, is a problem not unfamiliar to the professional, the
problem of conveying valid technical concepts to the layman. But the problem
is not a reason for avoiding the issue. The client who is well informed and who
knows both that he needs to understand uncertainty and that the valuer can
28 assist will no doubt wish to view and discuss the full range of sensitivity
calculations. Most clients will wish that there were no uncertainty and will
require support.
The solution must lie in the creation of some format description, accepted as
a norm, which conveys the essence with simplicity, but is capable of expansion
and interpretation. This would need to be presented in a prescribed
professional standard, and would always be appended to a valuation figure.
The valuer would then be required to enlarge on this format in the text of the
valuation report in the light of his/her knowledge of the client and purpose of
the valuation.
We would argue that there are six items of information which must be
conveyed. Using Figure 6, these are:
(1) the single figure valuation [V]
(2) the range of the most likely observation [V1-V2]
(3) the probability of the most likely observation [P3]
(4) the range of higher probability [Za-Zz]
(5) the range of 100 per cent probability [Va-Vz]
(6) the skewness of probabilities [Va -V1, V2-Vz]

Figure 6.
The single figure valuation. The study and description of uncertainty are not an Academic papers:
exercise in relieving the valuer of the need for expert decision. Clients usually Uncertainty in
need and will always expect ``a valuation'' to produce a single figure. property
The range of the most likely observation. While the valuer should be left free
valuation
to set the range and probability of observations in the manner most apt for the
task in hand, there is a strong case for prescribing some limit to the range of the
``most likely'' observation, 2.5 per cent or 5 per cent either side of the single 29
figure.
This would limit the fuzziness around the figure of greatest interest to the
client. A valuer who attached a 30 per cent aura of equal probability around
the central valuation figure would scarcely be helping the client! The
disadvantage of such a rule would be that, for some valuations, a narrow
range might be unrealistic. This ``rule'' would not apply if the valuer had
identified ``abnormal'' uncertainty. In that case he or she must have complete
freedom, with the concomitant obligation of explanation. It might follow
from this that if a valuer has difficulty in meeting the rule it may prove a
sign of ``abnormal uncertainty''.
The probability of the most likely observation. This needs no further
explanation. However, it is worth commenting that it might have a rather lower
rating than some might wish.
The range of higher probability. Even in a minimum description some feel
must be given for the flatness of the ladder. This is best achieved by giving
the range of the ``higher'' probabilities, coupled with the 100 per cent range.
The difficulty comes in defining ``higher''. This might be achieved by
defining it as the level above which more than, say, 60 per cent of all values
will lie.
The range of 100 per cent probability. The whole range of probability should
be given.
The skew of probabilities. It would be very common for the ladder not to have
an equal shape either side of V. If that is so then some feel for this shape must
be given as the skewness will be important information. To describe it fully
could, however, be laborious. The most practical solution would be to give the
sum of the probabilities applying to figures below the lowest figure of the ``most
likely'' observation, and above the highest figure. If skewness is of real interest
to the client considerable narrative description may also be necessary.
Taking the figures from Table II, the format might take the following form:
MV £ 2,575,000
Probabilities 2,500,000 ± 2,650,000 40 per cent
2,400,000 ± 2,750,000 75 per cent
2,025,000 ± 3,079,000 100 per cent
Skew less than 2,500,000 22 per cent
more than 2,650,000 33 per cent
JPIF This must be followed by some explanation. The degree of explanation would
18,1 reflect the valuer's knowledge of the client and the purpose of the valuation, i.e.
the importance of uncertainty in the particular case. Thus for an informed
client who was not intending to act upon the valuation, or for an uninformed
(and potentially confused!) client, a simple comment to the following effect may
be adequate:
30 I give, below my figure of valuation, a description of uncertainty required by the RICS
Manual. This is an entirely normal distribution for a property of this type and you should
treat my figure as accurate within the normal margin. However, please ask for any further
explanation you feel you need.

However, if the client is intending, or might reasonably be expected, to act on

the valuation further narrative would be essential, as it would be if the
distribution were not ``normal'', remembering that if it is truly abnormal. The
Manual already requires this to be reported.
The concept of ``normal distribution'' is important. The ladder shapes and
dimensions will, with experience, be found to fall into recurring patterns which
will be seen as ``normal''. Normality will vary, perhaps, from property type to
property type. The emergence of perhaps a number of normal patterns will
clarify understanding and simplify communication.

Abnormal uncertainty
This paper is concerned, first and foremost, with the normal uncertainty
which will occur in the valuer's mind when producing a valuation. The RICS
Appraisal and Valuation Manual recognises that, in some circumstances,
the degree of uncertainty will be ``abnormal'', in which case the valuer must
report the fact (Practice Statement 7.5.32) and may give the valuation in the
form of a range of values rather than a single figure. The Manual gives
examples of the circumstances in which abnormal uncertainty might arise,
e.g. legal difficulty with the property or financial market turmoil, but
gives no further guidance on either its identification or how it should be
reported.
The note above should greatly assist in both the identification and,
especially, the disciplined reporting of abnormal uncertainty. The graphical
shape of the ladder will, over time, acquire some essence of normality. It will be
found that certain parameters recur, not with arithmetic precision as they will
grow and reduce in different market and property conditions. If the valuer finds
that these common limits are being exceeded in a particular case it will be a
prompt that abnormality may exist and its cause should be identified.
Although the manual does not say so, abnormality must always have a cause
which is in itself ``abnormal'', and the valuer must be able to explain why it is
abnormal.
When it comes to reporting, it must be highly desirable that this should be
done within some common framework if clients are to be confident that valuers
are familiar with the feature (a feature which will worry the client) and have
proven techniques to manage and advise on it. If valuers have a common
technique for always reporting uncertainty it is a small step to extend that Academic papers:
technique to describe abnormality. Further, a robust technique for normality Uncertainty in
will give a benchmark against which abnormality may be compared. At property
present that would not be possible. valuation
Will the open recognition of uncertainty undermine the credibility
of valuers? 31
A valuation is a carefully produced judgement intended to carry the full weight
of the valuer's professional reputation. Will this be dissipated if he/she starts,
not implying, but declaring, uncertainty?
The objective is to inform a client about a valuation; it is not about
valuations generally. Therefore there should be no threat to the credibility of
the valuer or the valuation process; rather the reverse. If uncertainty truly
exists, and with the characteristics that we have described, then the single
figure valuation, standing alone, may too often be carrying a weight which it
cannot really bear. In recent years many commentators have referred to the
fallibility of property valuation. Even if one strips out the cases where the
valuation is demonstrably wrong, as is too frequently the case when matters
come before the courts, there are still concerns about fallibility. Valuers might
argue that clients' expectations may be too high, and their figure should not be
given the weight that it often is. If that is the case, then the recording of
uncertainty will provide the opportunity for getting the correct perspective.
The profession will be seen to be addressing and managing in the interest of
the client what is, to others, obvious.

Conclusion
In this paper we have discussed ``normal uncertainty'' in valuation. The
production of most valuations, and all property valuations, is a process
which involves managing probabilities; the valuer's task is to produce the
most probable answer. In the way that the results of this process are
currently presented, a single figure valuation with little explanation, and
perhaps even in the way in which valuations are produced, lie the seeds of
misunderstanding. Further, the uncertainties which lie in the valuer's mind
contain information which, for some clients in some circumstances, is of
great value. If the valuer is truly an expert, then understanding the full aura
of his/her considerations may be very important to a client who intends to
act on the valuation.
We have proposed one possible method by which uncertainty might be
measured and recorded. It is probable that other and better methods can be
devised which not only are intellectually superior, but also fit more closely
to the realities of the valuer's work. Indeed it may be that more than one
methodology will be required to cover the full range of valuation types and
JPIF circumstances. However, the advantage of the method proposed in this
18,1 paper is that it is not far from present techniques. Development can come
later if the concept is proven.
``Normal uncertainty'' is a universal and also an unsurprising fact of
property valuation. The open acknowledgement of that fact, and transparent
management of its implications, will enhance both the credibility and the
32 reputation of valuers.
The principal judgement of valuations must be of their utility, by clients.
The essence of my thesis is that this will be increased by the expression of
uncertainty. Beyond that principal judgement, values are also judged for their
accuracy by various observers, and by the courts for their probity.