Buy-to-let: bets with the housekeeping
Last week I posted an item about the implications of unsustainable house prices for owner occupiers. Buy-to-let investors merit a special entry.
10% of the UK housing stock is accounted for by the privately-rented sector. The buy-to-let fashion has not increased this proportion – only facilitated a shift from some pretty tough landlords, private but very professional, to a lot of amateurs. About 75% of privately-rented homes are owned by individuals with fewer than 5 properties. They are effectively running a business inside their household finances. It’s now a bad business to be in. They are also likely to be the wrong people to be running it.
Take the ‘Smiths’
About a year ago I produced a ‘money makeover’ for the FT, on behalf of a young couple with modest means and a pretty sound approach to money but one glaring inconsistency: a buy-to-let ‘portfolio’ of two London properties. Without meeting or speaking to them, the subjects of my makeover reminded me of ‘John and Sally Smith’, the fictional but typical financial services customers I chose to introduce and close my book, as a ‘before and after’ case history. They seemed sensible and well-organised people and obviously open to advice yet vulnerable to many of the popular myths surrounding investment. Were the real-life ‘Smiths’ really up for no-nonsense advice?
Prices in London having risen quite strongly since I participated in that makeover. The ‘Smiths’ could be forgiven for thinking this professional was not very smart. But the impact of high prices on both the housekeeping and long-term wealth creation is predictable, albeit quantitatively uncertain, and my advice is in no way diminished by even higher prices.
The ‘true’ cash flows, including financing costs, of a buy-to-let business are now likely to be negative or close to break-even. This conclusion may be reached either because purchases were made quite recently, in which case the yield on the purchase price is the relevant inflow and costs are the outflow, or because they bought much earlier, in which case the net yields on the greatly increased market value are not much higher than the opportunity cost of their risk-free return. Indeed, in London even the gross yields are likely to be below current mortgage rates and including other costs (after allowing for tax relief) the property is unlikely to be producing a positive cash flow – or ‘washing its face’.
The risks to finely-balanced cash flows (in business terms, the budget variance) are skewed to the downside, because of voids and bad debts. Both are likely to be causally linked to the two other key sources of variance: rental levels and interest rates. Why is it sensible to assume that the rent and interest rate contribution to the budget variance are also skewed to the downside?
House prices may be one of the things driving interest rates higher but rental growth is not what is driving house prices higher. In spite of all the bullish talk of shortages of homes, inward migration and City bonuses, rents are flat. They have failed to respond to the growth in the rest of the economy for too long to be a short-term blip. Indeed, the proximate effect of high house prices and fear of a crash has been to create new demand for renting from frustrated buyers. Without that, rents might actually have fallen.
What is going on here is not obviously consistent with the headline performance of the economy. But it is consistent with what has been happening to the tenants’ housekeeping money. Incomes have been rising less quickly than the prices of many non-discretionary spending items, which reflect higher energy, raw material and finance costs. The fuel for the housing boom has been money and credit. Not just because of the housing boom (though it is part of the problem for the Bank of England) but because of other cost pressures, this fuel supply is being gradually choked off.
This is old-fashioned talk, from when the business cycle could excite real fear. After a long period of lower (and less volatile) inflation, we have come to think that economic growth can be better managed and recessions avoided. It is certainly what Gordon Brown wants us to believe. But the risk is not just evidenced by the ‘path’, which does indeed appear more stable, but by the state that financial balances for the key sectors of the economy end up in (see an earlier article). These imply similar degrees of disequilibrium as in the bad old days of high inflation and cyclical volatility.
Since recession will first make sector balances worse (corporate profits fall, government tax revenues fall, household net savings are used to cushion weak incomes), there is also a possibility that interest rates will need to stay high for longer to support exchange rates in the countries with the greatest sector imbalances, which includes us.
In time, a cyclical downturn in final demand caused by tighter money and credit will allow interest rates to fall again. How quickly they fall and how much they fall also depends (in this particular cycle) on what happens to raw material prices. These are not simply a function of global economic demand. Other relevant angles this time are geo-political influences on energy supply and pricing and a massive margin of under-investment in base metals extraction. Along with the benefits to economic management from lower inflation, we have also enjoyed, as a raw materials consumer, a long winning streak over producer nations. This other deeply unfashionable economic concept, ‘the terms of trade’, has only reversed partially in the last few years and will not necessarily turn back in our favour quickly.
These are some good reasons for assuming the variance to buy-to-let cash flows is now heavily skewed to the downside.
Cash flows and the housekeeping
People like the Smiths can ill afford negative cash flows, because they impact their pension and family-raising objectives. Short-term ‘paper’ losses in capital values on long-term investments do not have this impact. They can affect confidence in continued saving but that is an emotional and voluntary reaction that flourishes largely where there is no financial plan as a context.
For the real-life Smiths in the makeover, with just two properties, I calculated that the variance over the next few years was between plus zero and minus £20,000 pa. As I say, like my John and Sally characters, the real-life ‘Smiths’ were sensible people and were adding to non-contractual savings at the rate of about £15,000 pa. This was one cushion against a cash drain that other buy-to-let amateurs might not have. But even for the ‘Smiths’ this was a cushion dependent on their household earnings holding up. Since ‘John’ was self-employed in an economy-sensitive sector, it was not a totally reliable cushion. ‘Sally’s’ income was more secure but on the other hand she wanted to stop full-time working as soon as possible to start a family.
What I suspect many buy-to-let landlords have in common is insufficient readily realisable cash or investments to meet periods of cash outflow, because as soon as there is enough spare cash to act as the equity in another highly-geared property, they tend to be back in the market. Even the ‘Smiths’ had less than a year of ‘worst-case outflows’ in cash.
My makeover subjects did not face personal bankruptcy. They were not behaving obviously irresponsibly. I did calculate the likelihood of their net equity being wiped out by a housing market correction as about one in four, but without any reason to expect problems servicing loans there was no reason to fear lenders foreclosing. Other buy-to-let investors may be more recent buyers than the ‘Smiths’ or more overstretched relative to their liquidity cushion and to their employment earnings prospects. In these households, negative cash flow could threaten their solvency.
Cash flows and the growth tradeoff
I could only speculate (without the opportunity to discuss it with the ‘Smiths’) that either they were underestimating the risks of their other objectives being set back or they thought the long-term growth in capital values would make the setbacks worth while. They needed to understand both to assess the rationality of their strategy.
As my other articles have explained, the long-term real growth prospects from housing investment at normal prices are much lower than people realise. If capital growth in normal conditions is as low as about 2% in real terms (adjusted for inflation), it requires high gross rental yields to justify the high budget variance as well as the variance in long-term real outcomes. With only two (or even five) properties, this required yield is arguably about 8% pa in money terms, compared with gross yields in London of 5% or less.
But we are not in normal conditions. With prices about 40% above their sustainable trend, real growth is more likely to be about 0-1% pa over the next decade or two, with an interim fall in real prices far more likely than the convenient scenario of levelling out. To derive a measure of expected real ‘total return’, we need to add to this capital growth a return contribution from net yields. For this exercise, which is not specific to how the property is financed, we can ignore financing costs. The key difference between gross and net expected yields is a provision not only for repairs and maintenance but actual improvement. The post-1957 data on which my growth projections are based are net of a steady improvement in the acceptable quality of housing due to additional investment. A property which is not maintained relative to the rest of the housing stock will not generate the same growth. It is like a company that pays high dividends but under-invests: it is in gradual wind-up, returning not just income but also your capital.
If you are struggling to follow this, will not Britain’s new amateur landlords? I have even over-simplified the issue here. Technically, we cannot calculate a total return by taking yield and adding growth when the reinvestment flows, which are irregular and lumpy, have to be made from yield. The correct calculation is an ‘internal rate of return’.
I usually suggest as a rule of thumb that the ‘true costs’ for repairs, maintenance and improvement are about 2% of the ‘normal’ capital value – equivalent to doubling the initial real investment over 40 years. However, I have seen estimates that the historical rate has been nearer double in 25 years. I would like to have more evidence for this.
My rule of thumb implies a simplified real total return from let housing in today’s market conditions of about 4% pa (5% gross less non financing costs of 2% leaves 3%, plus growth of 1% makes 4%).
Property versus other investments
Unlike equities, we cannot look at past buy-to-let investors and see what real total return they accepted, because of rent controls and tenancy protection. We have to look forwards at likely returns and sideways at what they compete with. Today, they compete with two investments at opposite poles of the ‘risk spectrum’.
Index linked gilts will secure the inflation protection you hope to get from property but with a government guarantee that there will be no real-return variance. What you get for that is about 1.4% real yield.
The risky alternative is equity investments. At No Monkey Business we estimate that a diversified mix of major equity markets will provide a real return of about 5% pa in real terms (compared with 6% in ‘normal’ market conditions). There is considerable uncertainty around this estimate but only one chance in ten that over a period as long as 20 years the worst-case outcome will be below the index linked gilt yield.
The 4% from property does not look so bad when compared with 5% from equities. But there are some big differences.
The equity return is evidenced by very long histories of income and capital growth. It is calculated after reinvestment decisions made by company management and fully disclosed in public accounts. And it is based on well-diversified exposure to hundreds of different companies.
The buy-to-let portfolio has nothing like the history of total returns to rely on. There is little evidence of how trade-offs between income and growth have been made by landlords in the past or how those trade-offs have affected the payoff, expressed as a real total return over long holding periods. And it is certainly not diversified.
For commercial property, it makes sense for prices to be set by ‘diversified investors’, like pension funds and life insurance companies. But these are not the people setting these very low yields in the housing market. Financial institutions remain largely unmoved by residential property which is why the rental proportion is still only about 10%.
A housekeeping risk premium
Finally, the likes of the ‘Smiths’ need an ‘additional’ excess return as the prospective payoff for running this business within their household budget, because it impacts their overall financial planning in a way that other investments, unless also financed by borrowings they take out, do not.