UnclePhilip wrote:I'm not at all sure if this is 'on topic' or this is the appropriate board; if not I'd be grateful to be directed to the right board.
Keeping enough money not invested in shares, so as to be able to have enough to last for say 3-4 years if dividends drop significantly and you'd not want to risk having to sell shares after a major market fall, seems sensible.
However, I do this with a simple savings account, and was told off recently for not using bonds for this instead.
As I've no idea about bonds for this purpose, I'd be grateful for advice, experiences etc to help me understand more....
On a terminology point, I would actually describe what you're calling a "cash buffer" as a "cash reserve". Both cash buffers and cash reserves exist, and they might be combined in a single cash account, but they serve somewhat different purposes. A cash buffer is there to smooth out reasonably predictable variations in cash receipts, such as the seasonal variations in dividend payments - in particular, with most of the large companies that HYPs invest in having financial years that end on December 31st or March 31st, and with many of them concentrating their dividend payout on their final dividends and paying those final dividends around 4-7 months after the ends of their financial years, there is a strong tendency for HYPers to receive very noticeably larger dividend payments in late spring, summer and early autumn than in the other half of the year. So a HYPer who is living off their HYP income and whose outgoings are fairly evenly distributed through the year is likely to want to be saving up their higher income during the summer months when dividends are plentiful in order to be able to draw on it during the leaner winter months. This is easily done with a cash account into which one pays all the HYP's dividends, and from which one makes a regular payment (probably monthly) of the averaged-out-over-a-year amount to one's current account. (There are details to be got right about that, such as making certain the cash buffer account doesn't go overdrawn even at its lowest point of the year, using the actual payment pattern of one's own HYP's companies, not the above generalisations about summer vs winter, and that it's a good idea to make the regular payments somewhat less than the averaged-out-over-a-year amount because a buffer account whose balance drifts upwards over the years produces fewer, less urgent problems that one whose balance drifts downwards and eventually goes overdrawn. But the above description is the essence of a cash buffer.)
All in all, therefore, a cash buffer is something which is expected to be used, in a pretty predictable way. Also, even if a HYP had the theoretically-possible (but never occurring in practice) payment pattern of all of its companies paying just one dividend per year and them all paying on the same date, the maximum size of cash buffer it would require is no more than a year's worth of the HYP's income; in practice, I think a couple of months' worth of the HYP's income is likely to be easily enough for the vast majority of HYPs.
On the other hand, a cash reserve is there to safeguard the HYPer's income against hard-to-predict events, such as the bouts of dividend-cutting that strike particular sectors from time to time (e.g. the dividend cuts by miners around the end of 2015 and 2016), or even companies pretty much across the entire economy occasionally (the financial crisis at the end of the 2000s and the recent Covid crisis are the two examples so far this century). Because of the unpredictability of those events, it's not possible to do financial planning for it in anything like the same way as it is for a cash buffer.
The point of making that distinction between cash buffers and cash reserves is that because of the differences in when they're needed, the HYPer's investment requirements are pretty different for the two. The cash buffer will fairly constantly be getting relatively small amounts added to it and withdrawn from it: the HYPer's requirements for it are primarily for those additions and withdrawals to happen smoothly and as far as possible free of risks and costs, while investment returns are relatively unimportant because they're going to be tiny at best. Basically, a bank account that has no charges if kept in credit and pays no interest does the job very well - perhaps even better than one that pays ludicrously low interest rates like 0.05% if one has tax returns to fill in and values the simplicity of not having the additional income source more than a pound or two of extra income. (Indeed, one can use one's current account for the job, though that does have the disadvantage of missing out on the 'budgeting' aspect of the regular monthly transfer to one's current account. Or one can use the cash balance of one's broker account for the job, provided one takes care not to regard
all of that cash balance as 'available for investment'.)
On the other hand, the HYPer's requirements for a cash reserve are for it to be there, and there as completely as possible, even during a major economic crisis, because that's precisely when it's most likely to be needed! It should also be accessible quickly if and when it's needed: investments which mature years in the future and either allow no access before maturity or charge high early-withdrawal penalties are not suitable. (Unless one can set up a 'ladder' of such investments, such that there is always one that is maturing in the near future and taking cash out of them as they mature will be enough to do the income reserve's job of supplying the HYP's missing dividend income. Such arrangements are possible, but in the normal situation of not needing to draw on the income reserve, they do require the HYPer to be constantly finding the next 'rung' in the ladder to reinvest maturity proceeds in - HYPers who regard companies occasionally being taken over or returning cash that needs reinvesting by other corporate actions as a not-very-welcome chore aren't likely to like that!)
Another factor is that because the cash reserve is likely to be left untouched for many years at a time, and it may well be a relatively large amount (maybe 1-3 years of income needs compared with 1-3 months), getting a decent investment return on it is a lot more desirable than for the cash buffer.
A small amount of a HYPer's cash reserve - say the first month's worth - has to be accessible quite quickly. That rapid-access requirement and its relatively small amount means that it has pretty similar requirements to the cash buffer, and I'd be inclined to simply add it to one's cash buffer arrangements - e.g. set up the cash buffer account on the basis that its lowest balance during the year should be more than a month of income needs, rather than just positive.
For the rest of the HYPer's cash reserve, which will probably be the vast bulk of it, my assessment of various types of bonds' suitability for the job is:
* Premium Bonds: excellent access, very high chance of their full value being obtained, investment returns very low on average, no worse than zero, and highly variable. IMHO pretty hard to beat for say the next 11 months' worth of the cash reserve (after the first month's worth literally being in cash as mentioned above), beyond that, still a decent candidate but their low investment returns are becoming a more significant factor as the amount put into them increases. Note that although withdrawals are supposed to take considerably under a month, I'm allowing for a month because both the HYPer and NS&I might not manage to be quick as usual in the early stages of an economic crisis.
* Government bonds (gilts): access varies according to when it's wanted - if on maturity, excellent, with a very high chance of their nominal value being obtained (though whether that's the full value one paid for them depends on what one paid for them!); if before maturity, very good (but a bit less than excellent because it depends on the markets being open for business as usual), but a chance of not obtaining full nominal value. Short-dated gilts will probably obtain at least close to full value, as their market price will be dominated by their payment of nominal value on maturity; long-dated gilts may well show a large shortfall depending on future interest rate expectations. Note that it's how long-dated they are when the HYPer might need to draw down their part of the cash reserve than counts: if other parts of one's cash reserve can be expected to keep one going for the next three years, then a gilt with 3 years to go to maturity can be counted as a gilt held to maturity in one's cash reserve calculations, and one with 5 years to go can be counted as one that will (potentially) be sold with just 2 years to go. Judging the potential investment returns is not easy, and I'm not by any means the person to tell you how to do it, but on general principles I'd expect them to be a bit better than those for Premium Bonds. So IMHO a potential candidate for the later years of the potential use of one's cash reserve, for the HYPer who is willing to spend some time learning how to invest in them and (since they're investments that mature) to maintain a 'ladder' of them - see my comments on that above.
* Corporate bonds: there's IMHO far too high a chance of the companies that issue them getting into financial trouble (either real or perceived) in a crisis, so that if one uses them as part of one's income reserve, there's too high a chance of them not being there when wanted. Not saying anything against investing in them in general - I know too little about that - but they do seem a pretty bad mismatch to the job of an income reserve to me.
* Investment bonds from insurance companies: pretty much like corporate bonds - too high a chance of them not being there when wanted. Especially given the insurance companies use of 'Market Value Adjustments' to say in effect "You know those returns we're indicating - well, we don't really mean them in the event of a crisis...".
On the subject of how much of a cash reserve I would want, there are at least a couple of points I would take into account. The first is how big a financial disaster to plan to be able to be able to survive. The correct answer to that is
not "any disaster, however big", for the simple reason that
no financial plan is proof against a sufficiently major combination of income-generating investments ceasing to generate income, annuity providers going bust, cash savings being destroyed by hyperinflation, government schemes collapsing under the strain, etc. You may
want (and probably do) to be able to survive any financial disaster, however great, but it's not a realistic financial planning aim.
So basically, I would 'stress test' a cash reserve by imagining a really challenging pattern of dividend cuts, probably one which is worse than both the financial crisis of about a dozen years ago and the Covid crisis as we've seen it so far but not hugely worse than them, and seeing how big a cash reserve you would need to get through it. For example, dividends cut overall by 60% in year 1 (I think the worst I've seen reported is TJH's ~50% in the financial crisis), no recovery in year 2 (I think recovery was observed in year 2 of the financial crisis, and some has been in year 2 of the Covid crisis), then recovery to 50% down, 40% down, 30% down, 20% down and 10% down in years 3, 4, 5, 6, 7 respectively (which I think is quite a bit slower than the recovery from the financial crisis - we haven't of course seen any of those years yet for the Covid crisis) before finally getting on track again in year 8.
Then project the use of the cash reserve forward through such a crisis. A very simple projection of that cash reserve through such a crisis, assuming that your HYP starts off only just providing your income needs, says that you've used 0.6 years' worth of the cash reserve after year 1 to replace the missing dividends, 0.6+0.6 = 1.2 years' worth of it after year 2, 0.6+0.6+0.5 = 1.7 years' worth of it after year 3, etc, up to 0.6+0.6+0.5+0.4+0.3+0.2+0.1 = 2.7 years' worth of it after year 7, when subsequent income is on track and the cash reserve stops draining away. If you started with 3 years' worth of cash reserve, you'll be OK but left feeling that it was quite a close-run thing! And there is the issue that although the cash reserve has stopped draining away, it's not replenishing. So you're left rather vulnerable to the
next crisis... Of course, it might happen that the HYP starts performing better than planned and starts replenishing the cash reserve, but any such outperformance is probably not going to be as spectacular as the crisis underperformance, so even if it does happen, you're likely to be left vulnerable to the next crisis for quite a lot of years.
That brings me on to my second point, which is that it's a very good idea not just to have a cash reserve, but also an income safety margin: an excess of the HYP's income over the income you need. That improves the situation in several respects - for example, suppose you start with a 10% safety margin, i.e. each year, the HYP delivers 1.1 times your income needs. Then that same 'stress test' pattern of dividend cuts delivers 0.44, 0.44, 0.55, 0.66, 0.77, 0.88 and 0.99 times your income needs in years 1 to 7 respectively, and so it depletes the cash reserve by 0.56 years of your income needs by the end of year 1, 0.56+0.56 = 1.12 years of them by the end of year 2, 0.56+0.56+0.45 = 1.57 years of them by the end of year 3, etc, up to 0.56+0.56+0.45+0.34+0.23+0.12+0.01 = 2.27 years of them by the end of year 7. And from year 8 onwards, the cash reserve starts replenishing at 0.10 years of income needs per year, without any need for the HYP to perform better than planned. That's not a fast replenishment rate, but it's better than nothing...
That's two benefits: one needs 2.7-2.27 = 0.43 years of income needs less in the cash reserve, and the plan allows for replenishing the cash reserve after the crisis is over. A third is that it takes more minor bursts of dividend-cutting in its stride or very nearly so: if e.g. your HYP has two companies in the same sector that between them supply 10% of the HYP's income, and that sector suffers problems that cause them to cancel their dividends in year 1, return with 60% of their previously-expected dividends in year 2, and fully return to their previously expected dividends from year 3 onwards, then without a safety margin you get 0.90 and 0.96 years of income needs in the two years of that mini-crisis, your cash reserve is depleted by 2-(0.90+0.96) = 0.14 years of income needs, and it won't be replenished unless the HYP performs better than expected. With the 10% safety margin, that becomes 0.99 and 1.056 years of income needs, and your cash reserve is barely touched in year 1 and that tiny depletion is replenished entirely in just a few months of year 2.
Those benefits do of course have a cost: all else being equal, getting a HYP that delivers 10% more income than you need requires 10% more to be invested in the HYP. As a rough back-of-the-envelope calculation, that's about 2 years of the income being supplied by the HYP, which from a position of having no safety margin means about 2 years of your income needs. So if you're starting from a position of having a HYP that just barely supplies your income needs plus a cash reserve of 3 years of your income needs, you should be able to shift to having a HYP with a 10% income safety margin, but expect your cash reserve to drop to about 1 year of income needs. That will still give you quite a lot of protection against bouts of dividend-cutting by companies, though not enough to withstand the most severe (and hopefully highly unlikely) 'stress test' bouts, and it will sail through minor bouts of dividend-cutting and have far fewer lower permanently-depleted-cash-reserve problems than the no-safety-margin, 3-year-cash-reserve alternative. Whether you feel that's a good bargain is up to you... (I personally probably would, though I might well actually prefer a 5% safety margin and a ~2-year cash reserve to both. Though in the situation of being faced with that choice, I would only retire if forced to, because spending another year building up to a 10% safety margin and a ~2-year cash reserve would be a substantial improvement on both IMHO. But all of that is of course just my personal attitude to the risks involved - others' feelings on the matter can quite legitimately differ, and probably will!)
An important wording point about the above that I would ask people to take into account before replying: it talks about income
needed, not income
wanted. If e.g. one has a HYP that one expects to produce £25k dividend income per year (and no other income sources, to simplify the example at the expense of its realism), and one also expects one's living expenses to be £25k per year, one might think that one has no income safety margin. But it's possible (and indeed probable IMHO) that some of those living expenses can be cut out in a crisis (and indeed it might not merely be possible but compulsory, as the examples of holidays and live entertainment in the Covid crisis show). If say £5k of those expected living expenses are ones that are wanted but can be cut out in a crisis, then one's income safety margin is the 25% excess of the £25k income over the £20k of essential living expenses.
Finally, I should say that I haven't taken various other factors into account that are likely to be significant in projecting the use of a cash reserve forward - for example, the effects of inflation. That isn't because they're unimportant, but just because they would greatly complicate this already long post. So don't treat any of the above as firm advice about how big a cash reserve to have - just as general guidance about how to go about making the decision, with you filling in extra details about what type(s) of crisis worry you.
Gengulphus