A property producing a 9% yield can be a worse investment than one producing 6%. Not because yield does not matter, and not because the 9% is necessarily exaggerated. It can genuinely produce 9% and still be the worse decision. The problem is that yield tells you what happens in one part of the investment. It tells you very little about the range of things that can happen after you buy it.
Expected value
I think about property using something I originally learnt through poker: expected value.
I don’t mean sitting down and assigning a precise percentage probability to every possible outcome. Property does not lend itself to that level of precision, and pretending otherwise would probably make the analysis worse.
It is a habit of thought. What can realistically happen from here? How likely is each outcome? How much does it help or hurt me if it happens?
The objective is not to correctly predict which future will arrive. It is to make decisions where the range of plausible futures is favourable.
A property brochure normally shows you one future. Expected value makes you price the rest of them.
Property is unusually good at hiding uncertainty
Most property investment material is built around point estimates. The rent will be £1,400 a month. The yield is 8%. The property is being sold at a 10% discount. The area is undergoing regeneration. Capital growth is projected at 5% a year.
Some of those numbers may be perfectly reasonable. That is not really the problem.
The problem is what happens when an uncertain future is presented in enough detail that it starts to feel like a calculation rather than an assumption. There is a big difference between saying a property could achieve £1,400 a month and putting £1,400 into a five-year spreadsheet. Once it is in the spreadsheet, every other calculation starts treating it as though it has already happened.
The same is true of growth, occupancy, service charges, refinance values and future demand.
This is why the 9% versus 6% comparison matters.
Imagine the 9% property genuinely produces its advertised income. It is cheap, cash generative and the rental demand exists. But the location has seen little capital appreciation for a long time. The buyer pool is mainly other investors. Lender appetite is narrower. If mortgage criteria tighten, your refinance options reduce. If investor sentiment changes, the resale market becomes thinner.
Now take the 6% property. The income is less exciting, but the stock is straightforward. Mainstream lenders are comfortable with it. There is owner-occupier demand as well as investor demand. The location has deeper employment and rental markets. If your plans change, you have several credible ways out.
The 9% property wins if the only question is which one produces more income today. That is not the only future you are buying.
Neither property is automatically wrong. The question is whether the extra three points are paying you for something you understand and can live with, or for the possibility of being stuck.
Property marketing naturally pushes in the opposite direction. A distribution of outcomes is difficult to put on a brochure. One clean number is easy. A page headed “8.7% yield” is considerably more persuasive than a page explaining the circumstances under which the asset could become difficult to refinance in seven years.
That does not require anybody to be lying. It is simply a very different way of framing the decision.
The repetitions problem
Poker made expected value intuitive to me because poker gives you repetitions.
You make thousands of decisions and the feedback arrives quickly. You can get your money in as an 80% favourite and lose, while knowing the loss does not make the original decision bad. You can also make an awful decision, get lucky and win. The result does not suddenly make the decision intelligent.
Do it often enough and the distinction between process and outcome becomes impossible to ignore.
Property gives you almost none of that.
The English Private Landlord Survey 2024 found that 45% of landlords in its sample owned only one rental property, while another 38% owned between two and four. In other words, 83% owned four or fewer. The figures relate to landlords who had registered deposits directly with tenancy deposit schemes, so they are not necessarily representative of the entire sector, but the scale of the problem is clear: most individual landlords are not making hundreds of acquisition decisions. They are making a handful.
That creates several problems.
The first is simply sample size. Most investors will make fewer acquisition decisions in a lifetime than a poker player makes in an hour.
The second is the lag. You can buy a property today and not really know what you bought for another decade. It might take years for a weak block to develop lender problems, for service charges to become painful, for a location to stagnate relative to stronger markets or for the resale market to reveal itself.
The third is noise. Suppose you bought in 2013 and sold in 2021 for considerably more money. How much of that result came from brilliant property selection, and how much came from falling borrowing costs, rising asset prices, constrained housing supply and a supportive market?
It is very difficult to know.
Then there is outcome bias. Human beings are good at constructing an explanation after we already know what happened. If something performs well, we remember why we liked it. If it performs badly, it becomes easier to remember the external factors nobody could have predicted.
Property makes this worse because the feedback takes so long that the original reasoning may barely be remembered by the time the outcome arrives.
And even the repetitions you do get are not necessarily separate. Someone may own eight properties and think they have eight pieces of evidence that their strategy works. But if all eight are similar properties, in similar locations, financed through similar lenders and bought during the same broad market cycle, they may simply have made the same bet eight times.
Property doesn’t just give you very few repetitions. A lot of the repetitions it does give you aren’t genuinely independent.
Experience is not the same as calibration
None of this means experience is worthless. Quite the opposite.
Experience teaches you what buildings tend to cause trouble, what refurbishments actually cost, which streets work, how lenders react, what tenants want and hundreds of other things no spreadsheet can teach you.
But experience alone cannot solve the repetitions problem.
Thirty years as a landlord and six purchases is still six acquisition decisions.
Someone can therefore be genuinely experienced and still be badly calibrated when it comes to choosing investments. That isn’t an insult. The constraint is arithmetic rather than effort.
A rising market can make weak decisions look intelligent for years. A difficult market can make perfectly sensible decisions look foolish for a while. If the only test of your method is whether the last property made money, you are always in danger of learning the wrong lesson.
So you need something before the outcome exists
If you cannot rely on thousands of repetitions to teach you which decisions are good, you need another form of discipline.
That is why I use a framework.
Not because I believe property can be reduced to a neat mathematical formula. It cannot. The value is in forcing the questions to happen in the right order, before you know what answer you want.
I start with mortgageability and exit liquidity. Can I finance this sensibly today, and is there a broad enough market of people who can finance it when I eventually want to leave?
Then the stock itself. Is this likely to remain a sensible, durable asset, or are there structural, leasehold, construction, maintenance or block-level issues likely to become more painful with time?
Then the location. Not whether it has a good regeneration story, but whether there is deep enough tenant and buyer demand for it to keep functioning when conditions are less favourable.
Only after that do I really want to talk about the return drivers: the credible upside and what is genuinely left of the income after realistic costs and friction.
Weakness is allowed. Some very good investments have obvious weaknesses. The question is whether you are being properly compensated for them.
High yield can compensate for inconvenience, ugliness, management burden or a problem you understand and can fix. It does not automatically compensate you for becoming trapped in an asset that future buyers cannot finance.
I also like investments that leave the owner with options. Hold it. Refinance it. Change how it is rented. Sell it to another landlord. Sell it to an owner-occupier. Release capital and use it somewhere else.
You may never use most of those options. They still have value.
A property that only works if one particular version of the future arrives needs to be exceptionally rewarding to justify that fragility.
The same problem applies to me
There is an obvious challenge to everything I have just written.
I do not have thousands of clean property repetitions either.
I have seen more deals, lender decisions, valuations, blocks, rental markets and client situations than the average individual investor, and that pattern recognition matters. But it still does not give me a controlled experiment.
That is precisely why I don’t think “trust my experience” is a sufficient investment method. I want something capable of disagreeing with me.
I want the questions written down before I become attached to the answer. I want mortgageability considered before an exciting return distracts me from it. I want a weak exit to remain a weak exit even when everything else about the property is appealing.
The commercial structure matters for the same reason. My fee does not change depending on whether the conclusion is to proceed, renegotiate, pause or reject. The analysis needs to be allowed to produce “don’t buy it” without that becoming a bad commercial outcome for the person giving the advice.
There is one consequence of thinking this way that can feel uncomfortable.
A property I reject may go on to perform brilliantly. That does not automatically mean rejecting it was a mistake.
Likewise, a property I approve may disappoint. That does not automatically mean the original decision was poor.
The question is what could reasonably have been known at the time, what range of outcomes you were exposing yourself to, and whether you were being adequately rewarded for accepting them.
That is why the framework is designed to make most opportunities fail. Not because most properties will lose money, but because you do not need to own most properties.
You only need enough decisions where the odds, the structure and the options available to you are sufficiently good.
Property does not give you enough repetitions for the results to teach you quickly. So the work has to happen before the result exists.
The objective isn’t to own every property that might work.
It is to become very good at ignoring the ones you never needed to own.
The Framework — the Risk Filter and the Industry Decoder — is £450. lucasjames.co.uk/framework