The Simple Path to Wealth

[GeoffF100]

There has indeed been considerable literature on SWR, and the conclusions are much more aligned with the idea that 4% has risk, with potential for some that follow that percentage going bust.

More money chasing rewards = lowest common denominator effect. Over the time that the 4% SWR was measured it was

1. Based on a right-tail good case market/period (US)
2. Had far fewer 'financially comfortable' investors looking for a share of the pie.

Level down 1 to a broader average (global stock index, which the UK was more reflective of) and the SWR is lower. Also factor in that more individuals are 'financially comfortable' and the denominator is lowered. If most are OK with 2% then prices/valuations will typically be bid up to just providing 2% instead of the good-old-days 4%.

You have to factor in asset price valuations which are much higher than they were historically. QE has drastically reduced interest rates and bond yields. The price premium for lower rated bonds has reduced. Equity valuations have increased correspondingly. Share prices have reduced since the coronavirus hit, but that is in response to a deterioration in the fundamentals. We cannot expect the same returns going forward as we did in the past.

People like to be told what they want to hear, but that does not mean it will it happen. Anything that returns more than an annuity carries more risk than an annuity, That risk is real, whatever people want to believe.

[TopOfDaMornin]

Personally I am 74-17 yrs retired
30/65/5% portfolio-equities/bonds/cash-Global Equities Index Tracker Fund and a Global Bond Index Tracker Fund hedged to the Pound-2 funds only-simple,cheap and easy to understand
2 years living expenses in cash
3.8% withdrawal rate pa
Portfolio bigger than when I started out
That does it for me-at the moment!
xxd09

This is what I personally would define as a simple path to wealth........and peace of mind.

[1nvest]

You have to factor in asset price valuations which are much higher than they were historically. QE has drastically reduced interest rates and bond yields. The price premium for lower rated bonds has reduced. Equity valuations have increased correspondingly. Share prices have reduced since the coronavirus hit, but that is in response to a deterioration in the fundamentals. We cannot expect the same returns going forward as we did in the past.

The Trinity study history/years already includes such high (and low valuation, and everything between) periods. Gains arise out of price appreciation, income/dividends, volatility capture (such as yearly rebalancing stock and bonds back to target weighting (a form of 'trading')). The rewards from each broadly compare overall, if that were not the case all investors would focus into the consistent best. Most strive to diversify across those potential rewards (but not all, Options traders for instance might just target Vega). Higher valuations/lower yields/interest rates equates to higher volatility (a 0.1% move in interest rates when yields are down at 1% invokes a more extreme move than a 0.1% move in interest rates when yields are at 5%).

Volatility could spike one year - such as share prices crashing heavily, and thereafter price appreciation and income could do well. There's nothing assured about a current/present high levels of x, y or z meaning that the rewards over the next 10, 20 whatever years will be below average.

Mid 1930's had high PE, low interest rates, some balanced asset allocations saw above average real returns over the subsequent 15 years - nearly twice the broader average. In other cases low valuations have seen subsequent below average outcomes, valuations declined further.

Post 2008/9 financial crisis when long dated (20+ year) gilt yields had declined to relatively low levels had some saying that it was mad to buy/hold such bonds, "yields only had one way to go" (upwards in their minds, inducing price declines (losses)). 2010 and 2011 saw around +10% and +20% gains respectively. Clearly they were wrong. As subsequent low yield (losses expectancy) mental train persisted further - so long dated gilts saw yet further double digit gains in 2014 (+20%), 2016 (+15%) and even again in 2019 (+11%). And its not as though the years between were big down years, 2013 was by far the worst, with a -7% loss and in all the others there was still a mildly (low single digit) nominal gain. Compounding to a +100% total gain since 2009 (to end of 2019), relative to +26% increase in inflation over the same years.

Yes forward time could see actuals outside of historical extremes, a new extreme, but on the basis of what economic/social/political history actually occurred a new extreme would be exceptionally severe. Hope that doesn't occur, don't think it will occur, but as ever there are no guarantees. Even if it did occur, then you have to actually get to physically live through it and out to the other side - that in itself may have a low probability given the circumstances.

For one of the benchmarks I use I'm seeing figures of a 5.5% annualised real (after inflation) average across the 2000's, pretty much around the longer term average. 2010's has seen that rise to 8.5%, I wouldn't be at all surprised if the 2020's saw that remain up at around 9% real levels, but equally could be down at around the 5% longer term average, maybe even a relatively poor/bad 4% level.


Those are yearly (in £'s total return values for US stock (and gold)) real (after inflation) figures. Rightmost are each decades averages figures. You can extend that Callan periodic table (with caveats - such as accounting for being on a gold standard pre 1932, when money and gold were exchangeable at a fixed rate it made more sense to have someone pay you to store your gold - i.e. hold Treasury Bills that paid interest) right back to the early 1800's and the rightmost values follow similar patterns/values. The only single exception was the 1910's, where the rightmost figure was around 0% - a decade of the Great War and what might be considered as the end of the British Empire.

UK home, US stock, gold ... diversified across £, primary reserve currency (US$) and global currency, holding land, equities, commodity assets ... and I very much suspect that the end of 2020 will look back with a rightmost figure being in the 4% to 9% range for the liquid asset holdings for that decade. No idea as to what sequence of returns actually source that outcome, whether price appreciation or volatility will have been the predominant driver, and would be very surprised if it was down at the 0% to 2% level that some are suggesting is the new SWR 'norm'.

[GeoffF100]

The Accumulator does not believe in the 4% rule either:

https://monevator.com/why-the-4-rule-doesnt-work/

I am sufficiently well funded that none of this matters to me.

[1nvest]

UK data since 1920 for all-stock had a 100% success rate if you averaged in over two years (3 timepoints)



You have to go back to 1896 to see the 20% failure rate that Wade Pfau suggested, predominately a consequence of the Great War years (WW1) clustering of bad case outcomes.

Without averaging since 1920 there was one 30 year failure run that started 1969 ... ended at the 2008 financial crisis lows. For similar reasons in the US, Japan's bad case outcome was in part a consequence of market cap weighting. Some stocks in similar sectors becoming giants that even when their value collapsed in half or more they still dominated the index. Midcap indexes feed in/out of both the top/bottom so are less inclined to see such domination of the index by single stocks/sectors.

Like you, I don't depend upon a 4% SWR, but for others it provides a guideline i.e. having to continue working/saving until having accumulated at least 25 times spending (so 4% SWR), but where that only might cover you for 30 years of retirement. A ballpark better than nothing guide. Many might equally not see even 10 years of retirement years. In some cases a individual might work to their 70's looking to get to having saved 33 or 50 times spending (3% or 2% SWR), and end up never having retired, and missed out on those retirement years compared to had they retired upon having saved 25x spending level.

Other options include buying 30 or whatever retirement years life expectancy you anticipate of liability matched inflation bonds (index linked gilts). Less of a viable option nowadays, but like everything, cyclical and in past years you could lock in at 3% real type levels, so for instance instead of having to load £20,000 into the 20th year maturity to have a inflation adjusted £20,000 spending amount in that year, you only had to load £11K present day value for £20K inflation adjusted value in 20 years time. As ever there are risks even with that, you could outlive the number or rungs (years) you loaded, or die before having spent all rungs, or the inflation rate for what you actually spent money on might differ from the broader consumer price inflation rate such that the intended liability matching didn't actually match your actual liabilities.

The safest approach is to accumulate way more than enough whilst still relatively young. Whilst you wont spend it, assuming appropriately invested to preserve the wealth, you wont go hungry. But for many that isn't a realistic option.

[GeoffF100]

You have to go back to 1896 to see the 20% failure rate that Wade Pfau suggested, predominately a consequence of the Great War years (WW1) clustering of bad case outcomes.

Clustering happens even with random data. Bad events have a tendency to precipitate other bad events. Selecting the most favourable period in the most favourable countries cannot be expected to be representative of the future.

Asset valuations are at an unprecedented high. The world appears to entering a less stable period. An epidemic that would kill about 1 in 200 if it was left to run free has been allowed to precipitate an economic disaster. Climate change is an existential threat, and is coming up inexorably, with no real action being taken. There are no great grounds for optimism.

A few decades ago few people considered anything other than an annuity to provide retirement income. Now they are going into draw down. They may be doing that at a very bad time.

[GeoffF100]

An important point here is that many people see the effect of high asset valuations on the income offered by an annuity, but they fail to see the likely effect of high asset valuations on the likely SWR for draw down. They just look at (or are shown) the SWR that could have been achieved historically when asset valuations were lower.

The financial services industry clearly does not have an interest in telling people that their SWR is likely to be poor. Neither does the government. "Your retirement income will be poor and we will do nothing about it" does not win elections. The FCA follows its political masters.

[dealtn]

There has indeed been considerable literature on SWR, and the conclusions are much more aligned with the idea that 4% has risk, with potential for some that follow that percentage going bust.

More money chasing rewards = lowest common denominator effect. Over the time that the 4% SWR was measured it was

1. Based on a right-tail good case market/period (US)
2. Had far fewer 'financially comfortable' investors looking for a share of the pie.

Level down 1 to a broader average (global stock index, which the UK was more reflective of) and the SWR is lower. Also factor in that more individuals are 'financially comfortable' and the denominator is lowered. If most are OK with 2% then prices/valuations will typically be bid up to just providing 2% instead of the good-old-days 4%.

You have to factor in asset price valuations which are much higher than they were historically. QE has drastically reduced interest rates and bond yields. The price premium for lower rated bonds has reduced. Equity valuations have increased correspondingly. Share prices have reduced since the coronavirus hit, but that is in response to a deterioration in the fundamentals. We cannot expect the same returns going forward as we did in the past.

People like to be told what they want to hear, but that does not mean it will it happen. Anything that returns more than an annuity carries more risk than an annuity, That risk is real, whatever people want to believe.

I think much of what you say here is true, and a better expression of what I believe your argument to be than your initial "If you draw 4% p.a. from the current market levels, you are very likely to go bust." If instead of a soundbite you had argued in such fashion initially I would have less concern.

I think the current market, both in terms of risk and valuation, are more associated with an unusual market than a normal one. As such you could certainly argue that risks are enhanced, and that relying on past studies of what a SWR is would be a riskier thing to do, and given that one of the outcomes of such a dependency might be bankruptcy it would be right to fully consider things.

That's still not enough to forecast "very likely to go bust". For one thing there is no compulsion to stick to a 4% rule should the likely negative consequences become more probable. However, as I think has been pointed out, the Trinity study itself encompassed many past periods which whilst different to that being seen today, could also be described as I have above in being closer to "unusual" than "normal", and so captures some of the uncertainties and difficulties (albeit perhaps different ones). Under that study, as has been pointed out very few "went bust" under a 25-30 year timescale, and beyond that the percentage falls considerably short of any meaningful definition of "very likely".

A strict Monte Carlo simulation of current market valuations and reasonable variations in returns, inflation, longevity (and any other variables) would I believe also fall a long way short of a demonstration of "very likely". Which isn't of course to prove bankruptcy wouldn't be possible, nor indeed unlikely for some, hence the extreme scepticism of anyone making claims about how easy it would be to either increase, or indeed maintain wealth. Certainly anyone making claims of how simple it would be to follow a rules based path and avoid the negative outcome tail-event of the probability distribution should be called out for it. But that process requires the symmetrical calling out of anyone making similar claims about how likely it is to result in such a tail event from following such a system.

[GeoffF100]

You have to factor in asset price valuations which are much higher than they were historically. QE has drastically reduced interest rates and bond yields. The price premium for lower rated bonds has reduced. Equity valuations have increased correspondingly. Share prices have reduced since the coronavirus hit, but that is in response to a deterioration in the fundamentals. We cannot expect the same returns going forward as we did in the past.

People like to be told what they want to hear, but that does not mean it will it happen. Anything that returns more than an annuity carries more risk than an annuity, That risk is real, whatever people want to believe.

I think much of what you say here is true, and a better expression of what I believe your argument to be than your initial "If you draw 4% p.a. from the current market levels, you are very likely to go bust." If instead of a soundbite you had argued in such fashion initially I would have less concern.

I think the current market, both in terms of risk and valuation, are more associated with an unusual market than a normal one. As such you could certainly argue that risks are enhanced, and that relying on past studies of what a SWR is would be a riskier thing to do, and given that one of the outcomes of such a dependency might be bankruptcy it would be right to fully consider things.

That's still not enough to forecast "very likely to go bust". For one thing there is no compulsion to stick to a 4% rule should the likely negative consequences become more probable. However, as I think has been pointed out, the Trinity study itself encompassed many past periods which whilst different to that being seen today, could also be described as I have above in being closer to "unusual" than "normal", and so captures some of the uncertainties and difficulties (albeit perhaps different ones). Under that study, as has been pointed out very few "went bust" under a 25-30 year timescale, and beyond that the percentage falls considerably short of any meaningful definition of "very likely".

A strict Monte Carlo simulation of current market valuations and reasonable variations in returns, inflation, longevity (and any other variables) would I believe also fall a long way short of a demonstration of "very likely". Which isn't of course to prove bankruptcy wouldn't be possible, nor indeed unlikely for some, hence the extreme scepticism of anyone making claims about how easy it would be to either increase, or indeed maintain wealth. Certainly anyone making claims of how simple it would be to follow a rules based path and avoid the negative outcome tail-event of the probability distribution should be called out for it. But that process requires the symmetrical calling out of anyone making similar claims about how likely it is to result in such a tail event from following such a system.

I was commenting on drawing down at 4% come what may, not drawing down at 4% initially, and then changing tack if things do not work out well. What is the probability threshold for "very likely to go bust"? 10% or more chance of failure I would suggest. Some of the studies quoted in the Monevator article considered that to be good.

[EthicsGradient]

I was commenting on drawing down at 4% come what may, not drawing down at 4% initially, and then changing tack if things do not work out well. What is the probability threshold for "very likely to go bust"? 10% or more chance of failure I would suggest. Some of the studies quoted in the Monevator article considered that to be good.

Ah, maybe it's a matter of terminology, then. "Very likely" is used by the IPCC, for instance, to mean "90% or more chance of happening". I think most people would see "very likely" as "over 50%", and probably well over 50%.

[dealtn]

What is the probability threshold for "very likely to go bust"? 10% or more chance of failure I would suggest.

Well that in the proverbial nutshell is the difference then! I wouldn't even consider that "likely", let alone "very likely".

I would suggest likely is >50% ie. more probable than not, so very likely would be at least 75% if not higher.

[GeoffF100]

What is the probability threshold for "very likely to go bust"? 10% or more chance of failure I would suggest.

Well that in the proverbial nutshell is the difference then! I wouldn't even consider that "likely", let alone "very likely".

I would suggest likely is >50% ie. more probable than not, so very likely would be at least 75% if not higher.

I found a variety of definitions for "likely". This one was the top Google hit:

"Such as well might happen or be true; probable."

https://www.lexico.com/en/definition/likely

Without a qualifier it means something might happen. I believe that "very likely" indicates that the probability is significant. In short, "likely" is a very vague term. I was deliberately vague. It is very difficult to give a meaningful probability value.

According to the Monevator article, the global portfolio had a SWR of 3.45% historically. Costs would have reduced that. High asset valuations are likely (again being deliberately vague) to reduce that. The present valuations are not high because the market thinks that there are excellent growth prospects (as would normally be the case). The high valuations are the result of bad things happening. Nonetheless, 4% could happen over the next 50 years, but I would not bet my money on it. For what it is worth, I think the probability of success is less than 50%.

[dealtn]

I found a variety of definitions for "likely". This one was the top Google hit:

"Such as well might happen or be true; probable."

https://www.lexico.com/en/definition/likely

Without a qualifier it means something might happen. I believe that "very likely" indicates that the probability is significant. In short, "likely" is a very vague term. I was deliberately vague. It is very difficult to give a meaningful probability value.

I'm not sure what I find more remarkable, your continuing belief on what likely means, let alone very likely, or the fact you taught statistics. Having a belief that anything other than something impossible to happen makes it likely.

I see another poster above concurs broadly with what I think likely and very likely mean in percentage terms. Others can draw their own conclusions I suppose.

Regardless, I think my conclusion is that anyone relying on a book alone to prescribe what to do hasn't done sufficient research, whether that is in acquiring, or maintaining "wealth".

[GeoffF100]

I found a variety of definitions for "likely". This one was the top Google hit:

"Such as well might happen or be true; probable."

https://www.lexico.com/en/definition/likely

Without a qualifier it means something might happen. I believe that "very likely" indicates that the probability is significant. In short, "likely" is a very vague term. I was deliberately vague. It is very difficult to give a meaningful probability value.

I'm not sure what I find more remarkable, your continuing belief on what likely means, let alone very likely, or the fact you taught statistics. Having a belief that anything other than something impossible to happen makes it likely.

I see another poster above concurs broadly with what I think likely and very likely mean in percentage terms. Others can draw their own conclusions I suppose.

Regardless, I think my conclusion is that anyone relying on a book alone to prescribe what to do hasn't done sufficient research, whether that is in acquiring, or maintaining "wealth".

You are just being rude again, and misrepresenting what I have said. If you want to know what someone means when they say that something is likely ask them.

I have said that I think that a 4% withdrawal rate has less than 50% chance of succeeding over the next 50 years. That is a little more precise, but still leaves a lot undefined. What I am saying is look before you jump. It is your responsibility to do your own research, not mine.

[1nvest]

Since 1825, history that includes some wild extremes of events, 50/50 UK stock/US stock fell short of a 30 year 4% annualised real gain in 10.9% of cases (yearly granularity). Dropping to 3% real and just 3.6% of cases fell short.



Fundamentally history suggests that wars = flat-line investment outcomes (or worse for the losing side). During major wars you might as well sell all liquid assets and just hold gold.

To me, the expectation of a 50% chance of falling short of a 30 year 4% annualised real is more towards a prediction of a prolonged major war, along with the loss/conquer of the US (home of capitalism). And where perhaps ones survival through such a event has high probability of failure. If you look at a example set of what year was the lowest withdrawal rate for a range of countries :

Denmark - 1939 (WW2)
South Africa - 1937 (WW2)
Netherlands - 1941 (WW2)
Sweden - 1914 (WW1)
Norway - 1912 (WW1)
Belgium - 1911 (WW1)
Finland - 1917 (WW1)
France - 1943 (WW2)
Italy - 1942 (WW2)
Germany - 1911 (WW1)
Austria - 1914 (WW1)
Japan - 1937 (WW2)

Possible yes, probable (50/50) IMO no, and even if so then likely we wouldn't be around to care anyway.

[GeoffF100]

Since 1825, history that includes some wild extremes of events, 50/50 UK stock/US stock fell short of a 30 year 4% annualised real gain in 10.9% of cases (yearly granularity). Dropping to 3% real and just 3.6% of cases fell short.



Fundamentally history suggests that wars = flat-line investment outcomes (or worse for the losing side). During major wars you might as well sell all liquid assets and just hold gold.

To me, the expectation of a 50% chance of falling short of a 30 year 4% annualised real is more towards a prediction of a prolonged major war, along with the loss/conquer of the US (home of capitalism). And where perhaps ones survival through such a event has high probability of failure. If you look at a example set of what year was the lowest withdrawal rate for a range of countries :

Denmark - 1939 (WW2)
South Africa - 1937 (WW2)
Netherlands - 1941 (WW2)
Sweden - 1914 (WW1)
Norway - 1912 (WW1)
Belgium - 1911 (WW1)
Finland - 1917 (WW1)
France - 1943 (WW2)
Italy - 1942 (WW2)
Germany - 1911 (WW1)
Austria - 1914 (WW1)
Japan - 1937 (WW2)

Possible yes, probable (50/50) IMO no, and even if so then likely we wouldn't be around to care anyway.

Even if there is a 30 year 4% annualised real gain, that does not mean that you would not go bust somewhere along the way if you always draw an index linked 4% from your portfolio. If you retire at 50 as I did, you may need to draw an income for 50 years, or possibly even more. Yes, you are most likely to go bust in very difficult times. Unfortunately, we can be sure that there will be difficult times ahead. That makes it more difficult, not easier.

[1nvest]

Even if there is a 30 year 4% annualised real gain, that does not mean that you would not go bust somewhere along the way if you always draw an index linked 4% from your portfolio. If you retire at 50 as I did, you may need to draw an income for 50 years, or possibly even more. Yes, you are most likely to go bust in very difficult times. Unfortunately, we can be sure that there will be difficult times ahead. That makes it more difficult, not easier.

I follow a Talmud type asset allocation, third each in land (home), stocks and gold. Retired in my mid 40's, now nearly 60. I've evolved into that Talmud choice over time, only been using it for the last decade or so.

SWR for me provides a steady inflation adjusted income, paid no matter what (taken out of total returns, whether up or down, unlike 'spending dividends' that can have high variability). If started today with a 0.2%/month SWR (2.4%/yearly) for all historic 50 year cases over the last 200 years that has sustained the original inflation adjusted start date capital value in the worst case, more often (median average) doubled up or more the capital value. i.e. had a high rate of being a perpetual withdrawal rate. Yes that 0.2%/month is only 2.4%/year so a relatively low SWR (greater potential of success). But that excluded imputed rent benefit i.e. is measured against house price only value, and where the 'inflation rate' was revised to include house price inflation (third house price increases, two thirds broad consumer price increases). Factor in that historic imputed rent benefit, the rent you'd otherwise have to find/pay if you were renting instead of owning, and that was 4.2% average, so 1.4% when proportioned to being a third of assets. Which combined with the 2.4% SWR = 3.8%. So in some respects I am drawing a near 4% SWR, and based on history that has a high probability of success. Nor do I see the prospect for that being poor/lower in forward time.

As rebalancing back towards around a third each in home/stock/gold is awkward with respect to the home value element, we use UK stock as a more liquid proxy for 'home' value (so 'home' = actual properties and UK stock, but predominately actual home values). For currency diversification UK home (£), US stock ($/primary reserve currency), global currency (gold). Asset diversity of land (home), stocks and commodity (gold). A reasonably balanced asset allocation. Yes each of the assets are volatile, but when it comes to drawing the monthly SWR often one is up more than the others (or down less), and taking from that helps towards rebalancing the assets back towards more equal weightings; And often is a form of taking profits off the table, and where those short term fast gains might later mean revert (disappear) such that it was better to have taken money off the table at that time.

Personally in that context I opine that 4% still has a good forward time prospect of success, but of course subject to choice of asset allocation/assets. A far better prospect of success over the next 30 years (or even 20 years) compared to my prospect of being around to see whether that was the case or not. I now have the added benefit/risk reduction that I'll be receiving a occupational pension soon, so much of past portfolio income will transition from being relied upon over to being more for luxuries/(adult) children. Even more so if/when I additionally receive the state pension on top of that (age 67/68). I suspect the combined pensions alone could provide sufficient income for needs, such that personal financial risk is very low.

For another with no occupational pension, that was renting instead of owning their own home, that was all-in on stock (such as HYP) and ignoring capital value/relying upon dividends to cover paying their rent and providing disposable income - for me that would be (very) high risk and far more inclined to lead to failure (too little in dividends at a time when share prices were low - double whammy type hit).

[bluedonkey]

Talmud Investing - will we now start to see another acronym for an investing style (TI), alongside TR, LTBH, HYP, etc?

[Itsallaguess]

Talmud Investing - will we now start to see another acronym for an investing style (TI), alongside TR, LTBH, HYP, etc?

We'll need a new board, of course...

Talmud Investing Theory - Practical

Cheers,

Itsallaguess

[OLTB]

Talmud Investing - will we now start to see another acronym for an investing style (TI), alongside TR, LTBH, HYP, etc?

We'll need a new board, of course...

Talmud Investing Theory - Practical

Cheers,

Itsallaguess

I can see it now - is Unilever really Talmud or should it be on the Talmud General board?

Cheers, OLTB.