Real estate investment appraisal b w amended (1)

TOPIC

WELCOME TO FINANCIAL MODELLING OF REAL ESTATE INVESTMENT ASSETS : AN INTRODUCTION TO THE BASICS

Before going on to do some financial modelling of potential real etate asset acquisitions, you might want to go through this basic introduction to the investment appraisal process.

FYI: This presentation is not meant to be printed out. You’ll get more out of it if you click through it.

Press F5 and Page Down

Topics covered

• Rationale, required knowledge and rough approach• A hypothetical example• The hypothetical cash flow• The missing detail• The Key Inputs

– Forecasting rents– Forecasting sale prices– Forecasting depreciation– Forecasting void costs– Setting the target rate of return

• Uncertainty

RATIONALE, REQUIRED KNOWLEDGE AND ROUGH APPROACH

What is the point of this stuff?

• BECAUSE MOST MAJOR INVESTORS UNDERTAKE FINANCIAL MODELLING OF REAL ESTATE ASSETS IN ORDER TO– TO WORK OUT HOW MUCH THEY CAN PAY

FOR AN ASSET THAT THEY ARE CONSIDERING BUYING

– TO WORK OUT WHETHER AN ASSET THAT THEY OWN WILL DELIVER THEIR REQUIRED RETURNS

Do I need to know anything first?• Yes - I’m afraid so• You need to know about basic financial maths

– About discounting and compounding– About discounted cash flows– About Net Present Value and Internal Rate of Return

• If in doubt, go over the Introduction to Financial Mathematics for Real Estate Appraisal

• You need to know some basic things about real estate e.g. how leases work and the meaning of some terms. It’s often difficult to know which terms need to be explained and which don’t. For instance, most people know what a lease is, but they may not necessarily know what a schedule of dilapidations involves.

Some basic concepts• AN ASSET’S VALUE IS TAKEN AS THE PRESENT VALUE OF ITS FUTURE

REVENUE STREAMS. • THIS IS A FUNCTION OF THREE THINGS

– THE REVENUE STREAM– THE TARGET RATE OF RETURN– TIME

• “INVESTMENT VALUE” IS THE WORTH OF AN ASSET TO AN INVESTOR OR A CLASS OF INVESTORS

• You can think of Investment Value as similar to equity analysts trying to establish whether companies’ share prices are overvalued or undervalued. Equity analysts often estimate the future cash flows of the companies to estimate the present value of the dividend flows. You see phrases like “intrinsic”, “fundamental” or “fair” value being used to mean the same thing as IV. They then compare this estimate of value to the actual price at which companies’ shares are trading .

A common approach is to

• Set out on an annual (but could be quarterly or monthly) basis– How much cash (do we estimate) will be paid out?– When (do we estimate that) it will be paid out?– How much cash (do we estimate) will be received? – When (do we estimate that) it will be received?

• You need to estimate a target rate of return• Then estimate the Gross Present Value of the

net cash flow produced by the asset

2. A HYPOTHETICAL EXAMPLE

Enough concepts, how about an example? Er, please.

• OK - let’s take a hypothetical example – and strip out some of the detail for now

• Where? A real estate asset in London’s West End• What? 10,000 sq m of office space• Who? It was let three years ago to a tenant on a 15

year lease with rent reviews to Market Rent every five years.

• How long? The investor has a seven year holding horizon

How much? Just the basic inputs for now• How much (1)? The tenant is paying £6,500,000 rent per annum

• How much (2)? Recent deals suggest that if let today the Market Rent is £7,500,000• How much (3)? The research department estimate that Market Rents will grow at

3.5% per annum for the next seven years.

• How much (4)? The research department also estimate that offices experience rental depreciation at 1% per annum – i.e. loses value

• How much (5)? The research department also think that the asset will sell for 20 times its rent in

seven year’s time – that’s an (exit) yield or cap rate of 5%• How much (6)? The investor has a target rate of return of 7.5% per

annum• How much (7)? The existing owner wants £150,000,000 for it.

We’re placing a lot of faith in our researchers!

3. THE CASH FLOW

Here goesYear Year 0 Year 1 Year 2 Year 3 Year 4 Year 5 Year 6 Year 7

This is the rent review five years into the lease. The lease started three years ago. The Market Rent is currently £7,500,000. However, the research department are forecasting Market Rents to grow at 3.5% per annum. However, this asset is also expected to depreciate at 1% per annum. £7,879,688 represents £7,500,000 grown at 2.5% per annum for two years. This stays fixed for five years until the next rent review.

Sale Price £178,302,863

The sale price is a product of the rent at sale and the rental multiplier (exit yield at sale). The rent at sale is expected to be £7,500,000 grown at 2.5% per annum for seven years. This is £8,915,143. The expected exit yield is 5% or the multiplier is 20. This gives £178,302,863.

PV @ 7.5% 1.0000 0.9302 0.8653 0.8050 0.7488 0.6966 0.6480 0.6028

GPV £146,731,169This the sum of the discounted cash flows.It is how much the investor should pay if they require a 7.5% return

Total revenues from rental income and sale price

(1+i)-n

PV factor * Net Cash Flow

Rental income £0 £6,500,000 £6,500,000 £7,879,688 £7,879,688 £7,879,688 £7,879,688 £7,879,688

Net cash flow £0 £6,500,000 £6,500,000 £7,879,688 £7,879,688 £7,879,688 £7,879,688 £186,182,551

DCF £0 £6,046,512 £5,624,662 £6,342,838 £5,900,314 £5,488,664 £5,105,734 £112,222,445

As you should realise, if they pay the £150 million, they

won’t receive their target rate of return of 7.5% per annum.

The IRR is below the target rate of return and the NPV is negative

Cost -£150,000,000 £0 £0 £0 £0 £0 £0 £0Rental income £0 £6,500,000 £6,500,000 £7,879,688 £7,879,688 £7,879,688 £7,879,688 £7,879,688Sale receipts £0 £0 £0 £0 £0 £0 £0 £178,302,863

Net cash flow -£150,000,000 £6,500,000 £6,500,000 £7,879,688 £7,879,688 £7,879,688 £7,879,688 £186,182,551

DCF -£150,000,000 £6,046,512 £5,624,662 £6,342,838 £5,900,314 £5,488,664 £5,105,734 £112,222,445

NPV at 7.5% -£3,268,831IRR 7.12%

Alternatively, if they are trying to work out how much to pay, there is no initial cost – and the surplus is what the asset is worth

to them

Year 0 1 2 3 4 5 6 7

Rental income £0 £6,500,000 £6,500,000 £7,879,688 £7,879,688 £7,879,688 £7,879,688 £7,879,688Sale receipts £0 £0 £0 £0 £0 £0 £0 £178,302,863

Net cash flow £0 £6,500,000 £6,500,000 £7,879,688 £7,879,688 £7,879,688 £7,879,688 £186,182,551

NPV at 7.5% £146,731,169

4. THE MISSING DETAIL

What detail was left out?Mainly costs and fees

Management costs Letting agents fees

Capital expenditure Selling agents fees

Void costs Buying agents fees

Rent review fees Legal fees

Stamp Duty (buying and selling)

For a large assets such as a shopping centre, the cash flow could contain dozens of rows. One for each individual tenant. I’ll expand the cash flow later in order to illustrate some of the issues.

THAT’S ALL THERE IS TO IT! I COULD HAVE ADDED A LOT OF IMPORTANT ‘CLUTTER’ BUT THE ESSENTIAL PRINCIPLES ARE HERE. SO, THE TECHNIQUE IS STRAIGHTFORWARD. IT IS THE INPUTS THAT ARE CRITICAL. LET’S LOOK THEM IN MORE DETAIL.

5. THE KEY INPUTS

5.1 FORECASTING RENTS

OK then, could you start with rental growth forecasts?• What do you want to know?• How are they produced? By who? What is being forecasted? How reliable are they?• You don’t ask for much. Well, the UK real estate market is one of the most forecasted real estate markets in the world. Forecasts are usually

produced by specialist forecasters who use econometric (look it up!) models to predict future rents. They tend to be produced by – major investors e.g. Prudential, Aviva, Standard Life etc. themselves,

– or by research departments of real estate services firms (JLL, DTZ, CBRE etc.),

– by some of the large investment banks (UBS, Morgan Stanley etc.)

– and by some specialist firms (e.g. PMA, IPD). • The Investment Property Forum take a quarterly survey of their forecasts of rental and capital growth in the IPD index.• So the IPF do a survey of forecasts of IPD! Are the forecasts any good?• Depends on what you mean by ‘good’? Not surprisingly, they disagree. If you take average of their forecasts, they tend to get the

direction of rental growth right. If they think rents will rise, they do tend to rise and similarly with falling…• Hmmm – so the numbers may not be reliable?• Nope – afraid not. Forecasters tend to claim that their main contribution is get the relativities right. Which sectors/cities will perform best

etc. it’s pretty impossible to get the numbers right.• But it seems to be locations not buildings that they’re forecasting?• Yes – they could be forecasting the rent of the average building in the average location or the best building in the best location or the typical

IPD building in a typical IPD location. It’s a bit like forecasting the weather. When the weather person predicts that it will be 10 degrees celsius in London, they’re not saying that every part of London will be that temperature. It could vary a lot according to local conditions.

• But they’re not forecasting the actual building being appraised?• Unlikely – they will leave it to the analyst to focus on the buildings and, possibly, to tweak the forecasts according to whether they think that

the building might be better or worse. • How do analysts or appraisers do that?• No idea!

5.2 FORECASTING SALE PRICES

Getting the future sale price must be even harder!

• Yep. Their forecasts of capital growth are really not good. Often, they get the direction of growth wrong?

• How do they manage that when they get the direction right with rents?• The real problem is forecasting yields. One forecaster said to me “To be able to forecast property

yields, I need to be able to forecast bond yields. If I could forecast bond yields, I wouldn’t be even trying to forecast property yields”

• You think that yields are inherently unforecastable?• Is that an actual word? Anyway, yes - it is difficult, er, extremely challenging.• But there’s lots of people forecasting share and bond markets • If someone’s prepared to pay for a forecast of something, then someone else will usually step forward

to sell them a forecast of something. • Cynical – even for one so old. I suppose that they’re not actually forecasting yields in the building

being appraised?• Right again. To be fair, real estate forecasters are happy to acknowledge the difficulty of forecasting

yields but, if you’re doing cash flow analyses, it has to be done because the forecasted sale price for commercial assets is a function of the forecasted rent and the forecasted yield at sale.

• What would you do?• Pardon

What would you do about the yield forecasting problem?

• I’m not a forecaster.• That’s not what I asked.• Well, ok, I’d look at where yields were in terms of their long term average. Let’s say London WE office

yields have averaged around 5% and rarely gone further below 4% or above 6% in the last two decades (which they haven’t). I’d anchor around that 5% if looking five years ahead to provide an indicator of the yield on the best/average office.

• But the office being appraised may have changed?• We can be pretty sure of two things about any property in, say, five years. Firstly, it will be five years older.• Thanks for that• You’re welcome…and, secondly, the unexpired lease term will very likely have changed for the worse. I

might then ‘tweak’ the yield to try to take this into account possibly in consultation with some of the investment agents who understand how buyers are, albeit currently, pricing these differences.

• Actually , it sounds fairly sensible.• It makes me nervous. It’s a bit ad hoc and atheoretical. The implied forecast is reversion to the mean for

the location yield and some tweaking to take into account changes to the building. I suppose that it is well-known that very simple forecasting models e.g. no change often beat sophisticated techniques in forecasting competitions. So- maybe there is some merit in a fairly simple approach. In a five year cash flow analysis, the exit yield is likely to be the most important single assumption. It is prone to a lot of uncertainty. This variable should really be subject to intense discussion and debate if it is the basis for bidding.

5.3 FORECASTING DEPRECIATION

What about depreciation? Where does that come from?

• Well, we know surprisingly little about it. We can be pretty sure that buildings lose value as they get older. This may be remedied by some regular or ‘lumpy’ capital expenditure to improve the asset.

• So it can involve some blend of spending money and losing value?• Yes. This needs to be incorporated into the cash flow. There’s some studies that suggest that rental

depreciation is about 1-2% per annum in offices, a bit higher for industrial and a bit lower for retail.• I would have thought that it is a bit like cars. You lose a lot of initially and then it starts to slow. You

probably only need to start spending money on the asset after five or six years as well.• I suspect that you’re right – but it doesn’t seem to have been really nailed by

researchers. In my experience, some analysts or appraisers just ignore it. I’m not sure but I don’t think that Argus (standard industry software for doing investment appraisals) has a field for it.

• It can’t be that important then. • Well, in a low growth era, it might be really important. If rents are growing at 20% on average, then

1% or 2% depreciation is probably trivial but if rents are growing at 1% or 2% per annum, then depreciation wipes it out.

• How would I find out about what is likely to be needed to be spent on a building and when?• This is probably within the domain of the asset and/or building managers. They tend to

have the most experience in maintaining, repairing and refurbishing assets.

5.4 FORECASTING VOID COSTS

What about things like voids?

• Well, tenants do vacate at the end of their leases, become insolvent and go out of business etc. If a building becomes empty, you could end up with a negative cash flow!

• What kind costs are there?• You end up with no income and having to pay to insure, secure, maintain and market the asset. You also become

responsible for the property taxes. If there’s a service charge for common areas…• How often do tenants leave? How long does it take to find a new one? What are the costs?• Don’t sound so stressed. I suspect that it varies a lot. IPD produce a lease events survey that tries to monitor these

types of things. JLL produce a measure of service charge costs (OSCAR). It tends to cover insurance, utilities, cleaning etc.

• What do I do in the cash flow when leases end? Tenants can either stay or go. I have to assume one or the other?• Well, you could do a number of things. If you have 10 leases expiring during the period of the investment appraisal,

then you could assume that some will stay and some will go. You can’t have much idea about which ones – they might not even know it themselves yet. But you can be pretty sure that some will go and that some will stay. Just pick some. Alternatively, you could incorporate some expected losses.

• How do I do that? • Well, it’s the estimated costs of vacation (lost income and costs) adjusted for the probability of incurring them. You

also have to make assumptions about the new lease and, albeit implicitly, the new tenants• This all sounds worryingly like shooting from the hip to me. The assumed numbers will never be accurate.• True. Neither will estimated rental growth, depreciation, exit yield, capex….• What’s the point then?• Pardon

5.4 SETTING THE TARGET RATE OF RETURN

I’ve often wondered how investors set a target rate of return for assets?

• Funny that – so have I. I think that it really depends on the investor and their particular targets and constraints. – Some investors are being benchmarked against an (IPD) index. They tend to set their target rate of return above

the forecasted (IPD) index returns.

– Other investors may be focussed on absolute return – they’ll pick a number (albeit with reasons).

– Others may try to build a risk premium (to add to bond yields with similar time horizons) for the asset looking at its tenants, location, leases, construction etc.

– Companies and REITs may be linking required returns to their weighted average cost of capital.• So it’s kind of horses for courses?• Afraid so. It’s not really a precision science. There are too many data uncertainties. The assumptions

embedded in standard finance pricing models such as the Capital Asset Pricing Model tend to be so divorced from real estate market reality that they are not capable of being applied.

• I’ve heard that there may be problems applying them to shares?• Yes – target rates of return…a bit like laws and sausages, you tend to lose respect for them when you

see how they’re made.• You could say the same thing about forecasts?• Indeed, you could• …and investment appraisals?• I’m not arguing

6 UNCERTAINTY

Why bother with this stuff at all?• My take on it is that it’s bit like what Churchill said about democracy being the worst possible

system of government…• Except for all the others.• There must be an alternative.• I’m struggling. You need a way of deciding whether you think a building will give you the

returns that – you estimate – you’ll need. These returns are in the future and the future has lots of Rumsfeldian known unknowns.

• You could rely some qualitative evaluation, intuition or gut feeling.• You could but intuition usually contains implied forecasts. You tend to end up putting

numbers on things eventually. It’s probably best to regard the cash flow analysis as an arena to provide a framework for decision-making. As long as users are aware that the numbers are uncertain, then it provides an imperfect but probably optimal basis for allocating capital. We tend to like everything to be neat and exact. We want to get ‘the answer’ and to avoid ambiguity. I think that you simply have to live with the ambiguity.

• What are the most important assumptions then?• I think that it’s depends on the combination of how much they affect the appraisal and how

uncertain they are. Some variables are really important but you can be pretty sure of getting close to the right answer Rent paid is clearly a fact. Market Rent - you should get close. Some are really uncertain but not really that important e.g. future sale costs but they’re only uncertain because future sale price is so uncertain. Go on - you guess.

How would you classify the following assumptions or variables?

• Rental growth? High or Low Importance? High or Low Certainty?• Rent passing? High or Low Importance? High or Low Certainty? • Depreciation? High or Low Importance? High or Low Certainty? • Exit yield? High or Low Importance? High or Low Certainty? • Target rate of return? High or Low Importance? High or Low Certainty? • Void costs? High or Low Importance? High or Low Certainty? • Market Rent? High or Low Importance? High or Low Certainty?

In hindsight, it was a badly worded question. Arguably all the variables are important but some are more difficult to estimate than others. You can be really sure about what the rent passing is and the target rate of return. You can be fairly sure about the level of Market Rent. But you can’t be sure about rental growth in the future and the yield it will sell for. The success of this appraisal depends on getting the latter right and these high importance/low certainty variables are then, arguably, the most important variables

OK. Let’s leave it there. We’ve covered a lot of ground. I’ll make a more detailed spreadsheet available so that

you can see how these things are done in Excel.