Musings on Cash Buffers

[TUK020]

Background
Much of the intent of an HYP strategy as originally laid out is to provide a high and growing income from a defined capital pot. As an alternative to an annuity or a index linked bond to provide for income in retirement.

Issue
The increased income available from directly held higher yield stocks comes at a price of increased volatility which is probably not welcome where the HYP forms a significant portion of the retirement income.
HYP1 will, over its two decades, have had instances of 40+% income drop approximately once per decade. Worse, there is a considerable element of sequence of returns risk here. HYP1 had its first major drop after 10 years of comfortable cruising. The drop could have been much more painful if it had occurred in the early stages.

Commonly quoted approach
Using a cash buffer to smooth out income when dividends drop, and rebuild the buffer in good times.

Comparing returns
If one's starting point to to get a reasonable level of dependability from an income stream, then oft compared returns are from thinks like B7 basket of investment trusts (thank you Luni) and HYP example portfolios. However all of these approaches have had revenue streams quoted and compared, and then talked about the possibility of using a buffer to smooth income.

In order to get like for like comparisons
It would make sense to model different cash buffer algorithms (* example one described below) to real world data on say B7 & HYP1 revenue variability , and possibly some form of statistical analysis to get a better grasp of sequence of returns risk. This a variation on the modelling done for Safe Withdrawal Rate (SWR).
Then one could pick one's 'dependability confidence level' for an investment strategy, and this would give the portion of initial capital that would need to be reserved in cash for one's buffer.
Once one has some idea of what an appropriate reserve is per investment strategy (e.g basket if ITs vs HYP stocks), then a more meaningful comparison of performance for the invested portion of the strategy.

Question to the longer term members of TLF & TMF: has any such analysis been done already? and where would I find it?

*Note 1
Sort of Buffer algorithm I am thinking of: Day 1: start with 18 months forecast dividend income in reserve buffer. Take 85% of annual Dividend income (averaged monthly?) as payout income, remainder accumulates in buffer. When buffer gets to 2 years give yourself a payrise.
If buffer between 18-24 months, pay freeze.
If buffer below 18 months (15% pay cut? seems reasonable to restrain spending if the income source seems troubled?)

Note 2
Someone will start quoting CTY as never having had a cut in ??? forever. Implied meaning that they don't need a buffer. I think it just means that we don't have data to make a meaningful assessment of the probability or severity of a cut in payout. Probably better to use a basket like B7.

[1nvest]

In order to get like for like comparisons

SWR is the easiest. Consistent inflation adjusted income - much more like a annuity, taken out of total return, and that can be used to compare entirely different assets/investments equally (doesn't matter if assets pay a high, low or no dividends).

Analysis wise there naturally are sources around such as https://earlyretirementnow.com/2019/02/ ... s-part-29/

[GoSeigen]

Background
Much of the intent of an HYP strategy as originally laid out is to provide a high and growing income from a defined capital pot. As an alternative to an annuity or a index linked bond to provide for income in retirement.

Issue
The increased income available from directly held higher yield stocks comes at a price of increased volatility which is probably not welcome where the HYP forms a significant portion of the retirement income.
HYP1 will, over its two decades, have had instances of 40+% income drop approximately once per decade. Worse, there is a considerable element of sequence of returns risk here. HYP1 had its first major drop after 10 years of comfortable cruising. The drop could have been much more painful if it had occurred in the early stages.

Commonly quoted approach
Using a cash buffer to smooth out income when dividends drop, and rebuild the buffer in good times.
[...etc...]

Hmmm.
-If we're talking about HYP here, then the assumption is that dividends don't get cut (or very seldom) but rather are steady and rising, or as put by one stalwart, dividends are much less volatile than capital. Diversification is supposed to deal with the rare cases where dividends are cut: the other dividends are rising so very soon soon the lost income is made up. If you are saying that dividend cutting is such a big problem that a new coping mechanism has to be introduced, are you not simply negating one of the key premises of the original argument, and thus calling into question the very basis of the HYP strategy?
-If we are talking about investment in general, outside of HYP-land, then why not simply treat cash as another asset, to be evaluated alongside the others, increasing and decreasing its allocation as conditions vary? This is how I deal with cash. Sometimes my levels are low, and other times much higher. If I need cash and don't have it, too bad, it is raised from the most cash-like securities or borrowing or something I think deserves to be sold for whatever reason.

Or perhaps there is something I have misunderstood.

GS

[GoSeigen]

-If we're talking about HYP here, then the assumption is that dividends don't get cut (or very seldom) but rather are steady and rising, or as put by one stalwart, dividends are much less volatile than capital. Diversification is supposed to deal with the rare cases where dividends are cut: the other dividends are rising so very soon soon the lost income is made up. If you are saying that dividend cutting is such a big problem that a new coping mechanism has to be introduced, are you not simply negating one of the key premises of the original argument, and thus calling into question the very basis of the HYP strategy?

The corollary, if it's not obvious, is that if the premise is correct, then of course a buffer is in most circumstances unnecessary, the original strategy being designed to cope with the expected minor fluctuations.

GS

[MrFoolish]

The corollary, if it's not obvious, is that if the premise is correct, then of course a buffer is in most circumstances unnecessary, the original strategy being designed to cope with the expected minor fluctuations.

Would you call the current fluctuations "minor"?

[Alaric]

The corollary, if it's not obvious, is that if the premise is correct, then of course a buffer is in most circumstances unnecessary, the original strategy being designed to cope with the expected minor fluctuations.

Diversification would seem intended to help deal with minor fluctuations. Whilst the risk of a 25% or 50% collapse in asset prices is handled by ignoring capital values, there's an assumption in the original writings that similar falls in dividend income just won't happen. Later when they did happen, it was a consequence of investing in equities and mitigation ideas were not suggested. I'm talking of articles published by TMF and perhaps other sites rather than comments made on discussion boards.

[GoSeigen]

The corollary, if it's not obvious, is that if the premise is correct, then of course a buffer is in most circumstances unnecessary, the original strategy being designed to cope with the expected minor fluctuations.

Would you call the current fluctuations "minor"?

Genuine question or rhetorical?

GS

[Arborbridge]

-If we're talking about HYP here, then the assumption is that dividends don't get cut (or very seldom) but rather are steady and rising, or as put by one stalwart, dividends are much less volatile than capital. Diversification is supposed to deal with the rare cases where dividends are cut: the other dividends are rising so very soon soon the lost income is made up. If you are saying that dividend cutting is such a big problem that a new coping mechanism has to be introduced, are you not simply negating one of the key premises of the original argument, and thus calling into question the very basis of the HYP strategy?


GS

That's not an assumption about HYP which I would choose to make. In fact, we know that in a severe crash (i.e. the credit crunch) dividends in general get cut, and even before that date, I don't remember anyone assuming that they would continue ever upward without a break.

I would say the consensus view was always that a) dividends are less volatile than share prices b) over a long period of time, dividends tend to rise (and share prices will follow). That's about it, I think - I doubt anyone believed that in a general market crash that dividends across the board would carry on growing.

Outside general economic contract or market crashes, it would be fair to say that HYP assumes that diversification helps to mitigate short term disasters from one sector or one group of companies. And I don't rememebr anyone saying whether dividend cuts would be rare or not! No one could predict that, and I'm darned sure that neither pyad nor his more enthusiastic followers ever said that.

So, no - I don't agree with your final sentence because it is based on a false argument - that is, the assumption in your first sentence quoted.

The common sense scheme of not drawing out all of one's income and to have an income reserve of some degree has been with HYPers since the earliest days of discussion it. Nothing new to see.

Arb.

[Arborbridge]

The corollary, if it's not obvious, is that if the premise is correct, then of course a buffer is in most circumstances unnecessary, the original strategy being designed to cope with the expected minor fluctuations.

Diversification would seem intended to help deal with minor fluctuations. Whilst the risk of a 25% or 50% collapse in asset prices is handled by ignoring capital values, there's an assumption in the original writings that similar falls in dividend income just won't happen. Later when they did happen, it was a consequence of investing in equities and mitigation ideas were not suggested. I'm talking of articles published by TMF and perhaps other sites rather than comments made on discussion boards.

I think the first part is correct. I'm not sure about the second part as regards articles, but certainly when it was clear that dividends could suffer roughly as much as share prices, mitigations were a hot topic on the discussion boards.

[MrFoolish]

Genuine question or rhetorical?
GS

Genuine question. I'm asking if you believe the current dividend falls are putting the HYP approach under strain?

[dealtn]

I would say the consensus view was always that a) dividends are less volatile than share prices

I am not sure I would recognise that description, nor consider that a consensus view.

By their nature Dividends are decided by Directors between 1 and 4 times a year, depending on your view and the company. By the very nature of them they are "smoothed" and less volatile, until the point where they aren't. Share Prices are made "daily" and will wander in a seemingly volatile, random way, but generally reflect the news about the company.

It is rare for a share price to "plummet" to zero overnight, and even with most "bad news" days falls are rarely of the magnitude of 25-50%. A dividend can be cut as far as zero "overnight".

So it may come down to what you consider to be a definition of "volatile" but your generic description doesn't capture adequately, at least to me, the volatility of Dividends.

[GoSeigen]

Genuine question or rhetorical?
GS

Genuine question. I'm asking if you believe the current dividend falls are putting the HYP approach under strain?

I think my earlier post answers that!

I've got this weird belief that no rigid investing strategy can work in all circumstances. HYP was a lovely idea, but at some point it's inflexibilities were going to be tested by a shock. I don't think it's for me alone to call the actual moment. I'm just tossing around alternative ways of viewing things.

There, I'm sitting on the fence -- what do you think? Or maybe answer on another thread cos I don't want to hijack the OP.

GS

[GoSeigen]

I would say the consensus view was always that a) dividends are less volatile than share prices

I am not sure I would recognise that description, nor consider that a consensus view.

By their nature Dividends are decided by Directors between 1 and 4 times a year, depending on your view and the company. By the very nature of them they are "smoothed" and less volatile, until the point where they aren't. Share Prices are made "daily" and will wander in a seemingly volatile, random way, but generally reflect the news about the company.

It is rare for a share price to "plummet" to zero overnight, and even with most "bad news" days falls are rarely of the magnitude of 25-50%. A dividend can be cut as far as zero "overnight".

So it may come down to what you consider to be a definition of "volatile" but your generic description doesn't capture adequately, at least to me, the volatility of Dividends.

Indeed, and shares have frequently been touted as superior (to keep it simple) to fixed interest because the dividends could be raised but that was rarely balanced by an honest recognition that they might also face dividend cuts and underperform the FI.

GS

[Arborbridge]

I would say the consensus view was always that a) dividends are less volatile than share prices

I am not sure I would recognise that description, nor consider that a consensus view.

Well, I guess that depends on who you ask. It's been commonly bandied about, let's just put it no more strongly, that share prices bounces around more than dividends - maybe that's just based on what may happen to a collection of shares rather than one particular share. Anyhow, whether it was the view amongst HYPers or general investors or not, is neither here nor there. The truth is that the credit crunch showed us just how much dividend income can fall, and from that I learnt a lesson, as I believe the HYPing community did too.

Arb.

[Arborbridge]

Genuine question or rhetorical?
GS

Genuine question. I'm asking if you believe the current dividend falls are putting the HYP approach under strain?

My answer would be: not as yet. We've been here before (2008) so I would - for the moment - say what has happened up to now, is within the bounds of normal for HYP. As this is the HYSS board, that comment presumably applies also to other forms of HY investing.


Arb.

[MrFoolish]

There, I'm sitting on the fence -- what do you think?

I think it's an incomplete strategy - suggested as a replacement for annuities, but seen to be rather lacking during a downturn. For a so-called simple strategy it seems to generate no end of debate and confusion. Yet we hear nothing from its founder on income reserves or other mitigations.

I also fail to see why international shares and ITs are so taboo. What is the benefit of being so UK-centric, especially with the all the unknowns of Brexit? And we have a distinct lack of forward looking tech shares in this country. For a long-term strategy, you don't want too high a concentration in old economy shares, IMHO. Go compare the performance of the FTSE with most other indices.

[Wizard]

There, I'm sitting on the fence -- what do you think?

I think it's an incomplete strategy - suggested as a replacement for annuities, but seen to be rather lacking during a downturn. For a so-called simple strategy it seems to generate no end of debate and confusion. Yet we hear nothing from its founder on income reserves or other mitigations.

I also fail to see why international shares and ITs are so taboo. What is the benefit of being so UK-centric, especially with the all the unknowns of Brexit? And we have a distinct lack of forward looking tech shares in this country. For a long-term strategy, you don't want too high a concentration in old economy shares, IMHO. Go compare the performance of the FTSE with most other indices.

PYAD's view was made clear not that long ago, I don't have time to search for the quote but it was something like... "...I've always said that if you can't deal with significant swings in income then HYP (and possibly equity investment in general) is not for you...". I took that to mean that unless you have a massive surplus of income over your needs don't try HYP.

Others have created cash buffers to try and address this, but if you are going to hold three years of income in cash that is a fair bit of capital presumably sitting in a deposit account earning next to nothing, which is a drag on returns. When HYP investors quote their yield on investment I suspect many do not factor in that cash buffer in their capital.

Oh and I completely agree on international shares, that is in my view the most unfathomable rule of all in HYPing.

[tjh290633]

Question to the longer term members of TLF & TMF: has any such analysis been done already? and where would I find it?

*Note 1
Sort of Buffer algorithm I am thinking of: Day 1: start with 18 months forecast dividend income in reserve buffer. Take 85% of annual Dividend income (averaged monthly?) as payout income, remainder accumulates in buffer. When buffer gets to 2 years give yourself a payrise.
If buffer between 18-24 months, pay freeze.
If buffer below 18 months (15% pay cut? seems reasonable to restrain spending if the income source seems troubled?)

Note 2
Someone will start quoting CTY as never having had a cut in ??? forever. Implied meaning that they don't need a buffer. I think it just means that we don't have data to make a meaningful assessment of the probability or severity of a cut in payout. Probably better to use a basket like B7.

I doubt that you will find ready answers. My own portfolio began with £2,400 in a PEP in 1987, when they first began. With retirement over 10 years away, my objective was to build a flow of dividend income to supplement whatever pensions I would end up with. The question of a buffer did not arise then, not has it since. I have always worked on the basis of having a healthy cash reserve, in case I needed a new car, the washing machine packed up, or we wanted to go on holiday. In terms of dividend income, my philosophy has aways been like the traditional response of Rolls Royce to questions about the power of their cars - adequate. My portfolio has seen some dividends withdrawn since 2007 for various reasons, but always less than the total received in teh year concerned. It has seen reinvestment every year, so the need for a discrete cash buffer has not arisen. I tend to look at dividends per income unit, rather than the absolute value of dividends. Over the years, that number has fallen in 8 years up to 5th April 2020. Those were 1989, 1993, 1994, 1996, 2003, 2004, 2010 and 2019. Cash dividends have fallen in 7 years, 1993, 1994, 1999, 2004, 2009, 2010, 2016 and, of course, the current incomplete tax year. Some of this is the result of special dividends being paid one year and not the next.

I suspect that there are very few HYPers who have started with a lump sum and never reinvested any of the accumulated dividends. Some may have followed this course, but not many.

My record of dividends per income unit is:

.            Inc Units
         .   Ordinary 
Year to      Divs/unit
05-Apr-88         2.83
05-Apr-89         2.25
05-Apr-90         3.40
05-Apr-91         4.67
05-Apr-92         5.94
05-Apr-93         5.52
05-Apr-94         5.31
05-Apr-95         6.45
05-Apr-96         6.27
05-Apr-97         7.13
05-Apr-98         7.55
05-Apr-99         7.92
05-Apr-00        10.79
05-Apr-01        11.39
05-Apr-02        12.46
05-Apr-03        11.68
05-Apr-04        11.13
05-Apr-05        13.03
05-Apr-06        14.21
05-Apr-07        15.18
05-Apr-08        18.73
05-Apr-09        21.60
05-Apr-10        11.91
05-Apr-11        15.12
05-Apr-12        17.78
05-Apr-13        19.93
05-Apr-14        20.34
05-Apr-15        21.35
05-Apr-16        21.68
05-Apr-17        24.17
05-Apr-18        27.02
05-Apr-19        26.36
05-Apr-20        29.71

TJH

[SalvorHardin]

I also fail to see why international shares and ITs are so taboo. What is the benefit of being so UK-centric, especially with the all the unknowns of Brexit? And we have a distinct lack of forward looking tech shares in this country. For a long-term strategy, you don't want too high a concentration in old economy shares, IMHO. Go compare the performance of the FTSE with most other indices.

It's "because Doris..."

1) HYPers can't have international shares because Doris didn't have any. Imagine Gollum saying "nasty foreign shares" and you get the idea.

A reasonable justification for nothing foreign was that it was harder to buy foreign shares and monitor them. That was true in Doris' era, but when the commercial internet emerged in the late 1990s it became much easier. Nowadays it's no more difficult to deal in and monitor companies listed in major overseas markets than in the UK.

I can vouch for the difficulty in buying and monitoring foreign shares back in the time of Doris, having bought my first American share in the mid 1980s (it was Genentech). Even getting a share price was difficult - I relied on checking someone's copy of the weekend Financial Times.

Back on TMF in the early 2000s there was a strong bias against foreign investments, even UK quoted companies whose businesses were wholly outside the UK. Particularly small cap. oils who were routinely described as "bongo tinpot".

2) HYP can't have investment trusts (or unit trusts) because Doris didn't have any ("nasty funds")

[Alaric]

I suspect that there are very few HYPers who have started with a lump sum and never reinvested any of the accumulated dividends. Some may have followed this course, but not many.

That is essentially following a growth strategy where the outcome is a full invested portfolio and cash is there on its merits or otherwise as an investment. Why the success or otherwise of such a strategy against any other wealth accumulation method isn't measured by the asset value so accumulated I find a mystery. There isn't going to be so much to reinvest in 2020-21 with a resulting loss of value, but that's a dent that could equally arise from an asset downturn which is periodic with equity investment.