I sold my rental property last year, after owning it over 20 years. It’s a lovely property, worth around £1m, right in the heart of London – near the middle of the map below. I used to live in it, I travel past it regularly, I know its neighbourhood well. The Modern Flat has genuinely been part of my life – in a way I can’t say for most assets I own.
As most readers would I think agree, I am a pretty numerate, analytical person. Yet looking back on the sale of the Modern Flat, in my decision to sell I made two stupid mistakes. I got two of the big numbers wrong. Not just a bit wrong, but properly, materially wrong.
There are lessons here about investing, about selling, and about property vs stocks/shares. Let’s take a look.
Mistake 1: A £200k, 33% mistake
How much is your house worth? This is a pretty fundamental question for literally hundreds of millions of people around the world. Here in the UK we are famously obsessed with property prices.
I can truthfully say I am not obsessed with property prices. But perhaps I should be more obsessed than I am.
I have loosely tracked the value of the Modern Flat over time. I certainly know what I paid for it – “around £400k” it the number in my head. This number, while not exactly right, was certainly close. I also know how big the property is – correctly (unlike The Proper House, doh!). So I know what I paid for it in £/sq ft terms.
The question of how much its value has risen over time is never an exact science. But in the case of the Modern Flat it isn’t too difficult. The flat hasn’t changed since I bought it. It is in a building with over 50 similar flats, which sell from time to time. Nearby there are at least 10 broadly similar buildings with broadly similar flats, which also sell from time to time. Our building has some unique advantages, but so do most buildings. So the process of estimating its value is not too complicated.
How to value your property
In England there are three data sources for estimating value:
Sold prices. Sale prices are recorded in a public register, Land Registry, 3 months after the transaction. This provides a useful dataset showing actual prices paid. Unlike some other countries, there are no ‘side deals’, so the price is what is paid. The limitation here is that you often may not have many data points – but in my flat’s case, there are over 10 transactions in the same building in the last 5-10 years.
Listed prices. Of course at any given moment there are far more flats listed for sale than actual transactions. And it is a doddle to use Rightmove or Zoopla to view them. Zoopla even shows the £/sq ft for you on many listings. The limitation here is that prices shown are ‘asking prices’ and, almost by definition, the longer a listing is viewable the lower the chance that the actual value is the asking price.
Estimated values. Zoopla broke the mould over 10 years ago by providing instant free valuations of any residential property. This allows you to estimate the value of any property – incuding your own – and see how it’s changed from month to month. The valuations are notoriously inaccurate, but they are something.
In any case, the number I had arrived at in my head was £1.2m – a capital gain of £800k. I knew this number was bit optimistic/speculative but I had a reasonable basis for thinking it.
The estate agents who I spoke to were, surprisingly, quite negative about the value. Only one thought I would get more than £1m for it, the others advised marketing for £1m or less. As it happens we chose one of the ‘realistic’ agents, not the ‘tell them what they want to hear’ agent.
What I ended up getting for the flat was almost exactly £1m. A gain of £600k, 33% less than I had been mentally budgeting.
My estimate that the flat was worth around £1.2m was based on, among other things, an almost identical flat in the building which sold in 2018 for £1m. But what I had misjudged was that firstly London property prices haven’t really increased since 2018, and secondly that covid-19 changed the game for London properties without any outdoor space – such as my flat. So I suspect I too could have sold my place for £1m in 2018, but I ended up instead selling it for £1m in 2024 – six years with only fairly modest rental income and no capital gains.
There are a few lessons here:
Get multiple data points and view points. The portals are invaluable here. But local agents have perspective too – their boots are on the ground every day – and talking to a few of them helped add perspective to what was happening in our building and nearby. Our building has a porter who was helpful too – though with hindsight he tended to tell me about the exceptions, not the rules.
Property valuations are highly uncertain. Even in my apartment block, a relatively liquid market by property standards, I was operating with a larger margin of error than I’d ever tolerate with publicly listed equities. Judgement is required, and luck matters. Maybe that flat which sold for £!m in 2018 got lucky – maybe that was really a mispriced sale, with the seller making a windfall gain.
Beware cognitive bias. This was my first stupid mistake. I was happy to see the £1m in 2018 data point, and less focused on the stories about London house prices flatlining over the last few years. Once I had £1.2m in my head I found arguments to support it, rather than remaining property objective. Luckily I wasn’t too prey to anchoring – once I heard that £1m might be the right number I pushed on (partly to get ahead of Labour’s October budget, with rumoured hikes in capital gains tax). I am fortunate that these numbers don’t change my life materially; but for many people selling a home is selling the majority of their net worth, and every £10k makes a difference – which makes cognitive bias much more of a risk.
Mistake 2: Underestimating my yields by 60%
After the property’s value (and thus capital gain), the other key number is the income it generates. In this respect, property is just like any other investment asset.
My Modern Flat generated almost £50k a year of rent. I thought of this, based on my erroneous £1.2m value figure, as a yield of just over 4%. It was actually a gross yield of about 5% – actually quite high for central London.
I’m not that stupid
I think I avoided the obvious stupid mistake which is to consider my gross yield as my rental income.
As part of my annual tax filing process, I have quite a clear spreadsheet which itemises the various expenses I have to deduct from my annual rent. First of all, there is the letting agent’s commission – something I could have avoided if I wanted to put more time into the property and was prepared to take on the risk of being non-compliant with ever-changing local and national regulations myself. Secondly was the service charge, which I had no control over. Thirdly there was the mortgage interest – on my fairly small mortgage. And then there were smaller numbers like landlord’s contents insurance, a few small call-out/etc charges, and the annual ground rent.
My actual net income was just over £25k, about 55% of my gross rent. Based, again, on my erroneous valuation estimate this was a yield of about 2%. In fact I now know it was a net yield of about 2.5%.
So, what was my stupid mistake?
My mistake was then in how I was thinking about my non-cash expense: capital gains tax. This was the tax that I was going to have to pay when I sold the property. The issue arises because of the way ‘accidental landlords’ like me, whose property used to be our principal residence, are treated. To simplify, the period of ownership which falls under principal residence is tax-free, and only the rental period is taxable.
Here is what I wrote about this CGT liability, over two years ago. At that point, I was mindful that I’d owned the place for over 20 years, and the rental net income was just over £20k per year:
Capital gains tax relief is withdrawn. Capital gains on rental properties, on eventual sale, will be taxed pro rata to the portion of ownership spent as a principal resident. This is a cashless tax charge, only paid on sale. But for my Modern Flat, it is quite a big number. Assuming my gain right now is around £800k, roughly speaking 5% more (c.£40k) of this gain becomes capital gains taxable (at 28%) every year. So I have a roughly £11k additional tax bill every year. That almost wipes out any economic surplus I make from my rental property.
My basic logic here was that having owned the place about 20 years, each extra year was about 5% more CGT liability. And based on my estimate of value, this meant extra taxable gain of about £40k per year.
I had been thinking that, after owning the flat for 20+ years, every extra year was (approximately) an extra 20th, i.e. 5%, or so of extra liability. You can see the logic in the quote shown above. At the higher property CGT rates of 28% (until April 2024, when they were cut to 24% – and not changed by Labour later that year despite many rumours to the contrary), and an (erroneous) unrealised gain of £800k, I (erroneously) thought my liability was increasing by 28%x5%x£800k i.e. £11k a year. This was a material sum in my calculations.
This mistake left me thinking that my actual, cash plus non-cash, net income from the property – assuming the property wasn’t changing in value (a correct assumption, as it turned out), was less than 1% of the value of the property. This strongly influenced my decision to sell the property.
Once I ended up calculating the exact liability, it turns out my annual increase in exposure was less than £1k a year. This only became clear once I did the exact calculation. And with hindsight, and a bit of algebra, I can see my mistake. The key fraction is the period of time that I had lived in the property – about 3 of the 20+ years – and this fraction wasn’t reducing by 5% a year after all. Put simply, if after 20 years the principal residence (i.e. tax-free) component was 3/20ths, i.e. 15%, of the gain then after 21 years this component was about 3/21ths, which is 14.3%. The taxable gain isn’t even increasing by £6k a year, let alone the liability at 24% of that.
Assuming no increase in the value of the property, my exposure was actually climbing by about £700 per year. And my net yield was around 2.5% of the value, not less than 1%. Which, I can argue, was a 60% error.
There are two key lessons I draw here:
Don’t rely on mental arithmetic. I was using the lazy approximation that every year the liability fraction increased by ‘1 year’s worth’. This wasn’t right. If I’d actually built the tiny spreadsheet for myself ahead of selling, I would have realised the number wasn’t what I expected.
Log your expected CGT liability annually. More to the point, though I have kept records of my taxable income every year, I have not recorded each year that capital gain side. If I’d logged, each year, my estimate of value, and my calculation of CGT if I sold at the end of the tax year, both numbers that would have taken me 5 minutes max to add to my spreadsheet, I’d have quickly noticed that my CGT liability wasn’t really changing at all from year to year. For my investment portfolio I have started logging each month my unrealised capital gains – so I have a record over time. Once set up, it doesn’t take any time at all.
Two stupid mistakes, one right decision
So, I made material mistakes on both the asset value and the asset income.
If I’d known the property had only gained 2/3rds what I thought it had gained, and was yielding about 2.5x more net income than I thought, would I still have sold it?
The clear answer here is Yes, I would still have sold the flat.
They say two wrongs don’t make a right. In this case my two wrongs didn’t really make a difference to my decision.
The property being worth less than I thought, if anything, strengthens my argument that it has proved to be a poor investment – especially since 2018.
The net return being somewhat higher than I thought still leaves the property yielding only 2.5%, and not gaining in value at all – so its total return is only 2.5%. Compare that to my average annual return on my invested portfolio of over 8%. Yes, I prefer 2.5% to 1%. But I much prefer 8% to 2.5%. Even the dividend / income yield from my invested portfolio is around 3%, i.e. higher than I have been receiving from the rental property.
In addition, in 2024 there was a lot of uncertainty created by politics and in particular the new Labour government. To summarise, ahead of the October 2024 budget everybody thought taxes on landlords were going up. The uncertainty of this created a strong reason to get a deal done before that budget if possible. As luck would have it, I had already started the marketing process of selling the flat before the July 2024 general election so I was in a good position to sell in time – which I managed to do.
So the other set of decisions I made – the tactical decisions on how to sell, with which agent, whether to accept the (two) offers I had for the property – I think I got about right.
The final lesson: stocks/shares are a lot easier, aren’t they? What would happen if I put the money from the flat into a portfolio of stocks/shares instead? Well, we’ll take a look at that in future blog posts.