Drawdown rules help

mrstout

Does anyone here understand the Guyton Klinger withdrawal strategy - I cannot post a link as a newbie but it is easily found on Google.

In short, the theory is that a higher withdrawal rate is sustainable if small adjustments are made according to portfolio performance. But I cannot see how the rules work based on the published explanations.

There are four rules, but I cannot make sense of the ones which relate to these statements:
Assets with positive returns, resulting in over-weighted allocation excess is sold and the proceeds "invested" in cash to meet "future" withdrawal requirements.

Assuming the starting target allocation is:
£50,000 - 50% equities
£40,000 - 40% bonds
£10,000 - 10% cash

After 1 year at rebalancing equities have gained and the portfolio is now:
£55,000 equities (52.4%) 2.4% overweight
£39,000 bonds (37.5%) -2.5% underweight
£10,100 cash (9.7%) -0.3% underweight
£104,100 total

Normally the rebalance would be:
Equities £55,000 sell £2,950 = £52,050 (50%)
Bonds £39,000 add (buy) £2640 = £41,640 (40%)
Cash £10,100 add £310 = £10,410 (10%)
Total £104,100

However the rules infer selling all over-weighted allocations and adding to cash, so:
Equities £55,000 sell £2,950 = £52,050 (50%)
Cash £10,100 add £2,950 = £13,050 (12.5%)
Bonds (no add or sell) £39,000 (37.5%)
Total £104,100

So the portfolio allocation is now on target for equities, underweight on bonds, overweight cash. Is this right? Surely the allocations can drift further and further if this is repeated annually?

Perhaps the cash is not just for "future" withdrawals as stated, but is to be taken as that years withdrawal during the drawdown stage?
If so at 4% withdrwal rate of the starting £100,000 portfolio, £4,000 (ignoring an inflation adjustment) is withdrawn from cash:
Equities £52,050 (52%)
Bonds £39,000 (39%)
Cash £9,050 (9%)
Total £100,100

Not too far away from target allocation but surely still not what is intended? In fact the decision rules go on to explain that "withdrawals are funded in the following order":
Overweigh equities
Overweight fixed income (bonds?)
Cash
Underweight fixed income assets in order of prior years performance (i.e. the least worst first)
Underweight equties in order of prior years performance (i.e. the least worst first)
No withdrawals from equities in a year of negative returns if cash or fixed income assets are sufficient to meet withdrawal amount.

So I have these questions:
When it says withdrawals are "funded", this means "assets are sold"?
When it says "No withdrawals from equities in a year of negative returns" does it mean if the portfolio total return was negative, or if equities held (the asset class) were negative.
If several equity asset classes are held and some were negative and some positive, then what?
If profitable assets (e.g. equities) are constantly sold and held as cash, after a few years of strong performance surely there is a risk of holding far too much cash?

Can anyone explain how the rules are supposed to work, possibly using my example allocations above to illustrate where my interpretation may be wrong?

bigadaj

You seem to be missing the drawdown element.

The whole point is to take an income from the investment as a predetermined but variable level.

So assuming say a 4% withdrawal rate then you take this unless the value of your investments fall, it's aimed at the decumulation stage of your life, not accumulation.

So withdrawals can be funded from income, capital appreciation or be reduced or suspended due to relative poor performance.

mrstout

bigadaj wrote: »

You seem to be missing the drawdown element.

The whole point is to take an income from the investment as a predetermined but variable level.

So assuming say a 4% withdrawal rate then you take this unless the value of your investments fall, it's aimed at the decumulation stage of your life, not accumulation.

So withdrawals can be funded from income, capital appreciation or be reduced or suspended due to relative poor performance.

I know that the rules relate to the drawdown stage. But if the excess weighting of the strongest performing asset is well over the 4% needed to fund the annual withdrawal, and the rule requires all the excess weighting is sold and the remainder (over the necessary 4%) is held as cash "to meet future withdrawal requirements", if this is repeated over several years could the portfolio not become far too overweight in cash?

jamesd

If you haven't yet read their article Decision Rules and Maximum Initial Withdrawal Rates that might help to understand how it is supposed to work. You'll probably also find the material in Drawdown: safe withdrawal rates of interest, particularly Guyton's sequence of return risk taming approach.

Here's a paraphrasing of the Guyton and Klinger rules:

1. Find out how much the portfolio increased or decreased over the past year. Also calculate what taking the just ending year's income would be as a percentage of the current portfolio/pot value.
2. If it's more than 15 years from the planned end of drawdown, is the just ending year's income level now more than 20% above the percentage it was when drawdown started? So if it was 5% at the start of drawdown, would taking the ending year's income be 6% or more of the current pot value? If it is, reduce the past income used in later calculations by 10%. This is the capital preservation rule.
3. Is the current percentage of income being taken more than 20% below the initial percentage? So if it was 5% initially, would it be 4% or lower now? If it would be, increase the past income used in later calculations by 10%. This is the prosperity rule.
4. Did the portfolio decrease in value? If it did, would taking the same income cause the percentage of the pot being taken to be higher than the initial percentage? If it would be higher, don't increase for inflation, else do. Don't do anything to catch up for years of no increase. This is the withdrawal rule.
5. How to fund it, the portfolio management rule:
A. If anything had a positive return and is now at an overweight allocation, sell the overweight amount and put that in cash.
B. If equities are overweight, sell enough of them to get back to target weight, putting the proceeds in cash.
C. If fixed income is overweight, sell enough of them to get back the target weight, putting the proceeds in cash.
D. If there's enough cash to pay the income now, use that.
E. If still short, sell fixed income investments to make up the shortfall.
F. If still short, sell enough in equities to make up the shortfall, starting with the ones that had the best performance in the last year first.
6. I you didn't rebalance fully in step 5 by selling, rebalance the portfolio now, after having taken the income.

jamesd

mrstout wrote: »

Assuming the starting target allocation is:
£50,000 - 50% equities
£40,000 - 40% bonds
£10,000 - 10% cash

After 1 year at rebalancing equities have gained and the portfolio is now:
£55,000 equities (52.4%) 2.4% overweight
£39,000 bonds (37.5%) -2.5% underweight
£10,100 cash (9.7%) -0.3% underweight
£104,100 total

A. Equities had a positive return and are 2.4% overweight, sell 2.4% and place the £2,519 in cash. Bonds didn't have a positive return, don't sell them.
B. Nothing to do, already sold the overweight equities because their return was positive.
C. Fixed income underweight, nothing to do.
D. If nil inflation and still 4% income target you need to raise £4,000 for income. £2,519 of this is funded from the equity selling cash.
E. Still short, sell £1,481 of bonds to cover it. That's the £4,000 funded so stop here.
F. If no bonds were left, sell equities in best past year performance to worst order.

mrstout wrote: »

So the portfolio allocation is now on target for equities, underweight on bonds, overweight cash. Is this right? Surely the allocations can drift further and further if this is repeated annually?

That's right. You could even run out of cash and bonds if equities have a long series of bad years.

This is deliberate, so you don't sell out of equities during a down period unless you have no choice, it's protecting you from selling low for as long as possible, so you can outlast sustained bad periods. Equities drop more than bonds generally so it's the equities that have to be protected from selling.

mrstout wrote: »

Perhaps the cash is not just for "future" withdrawals as stated, but is to be taken as that years withdrawal during the drawdown stage?

Yes, it is. Do the overweight selling first, then pay the income out of cash from past years plus that from overweight selling. If still not enough to pay out, sell bonds first, then equities if out of bonds.

mrstout wrote: »

When it says withdrawals are "funded", this means "assets are sold"?

Yes.

mrstout wrote: »

When it says "No withdrawals from equities in a year of negative returns" does it mean if the portfolio total return was negative, or if equities held (the asset class) were negative.

Just the equities asset class.

mrstout wrote: »

If several equity asset classes are held and some were negative and some positive, then what?

Start selling in best equity performance in the last year to worst performance in the last year order.

mrstout wrote: »

If profitable assets (e.g. equities) are constantly sold and held as cash, after a few years of strong performance surely there is a risk of holding far too much cash?

Sell high is a deliberate part of the plan. The prosperity rule will eventually be triggered to soak up the cash.

mrstout

Much clearer now thank you jamesd!

This bit of the example:

D. If nil inflation and still 4% income target you need to raise £4,000 for income. £2,519 of this is funded from the equity selling cash.
E. Still short, sell £1,481 of bonds to cover it. That's the £4,000 funded so stop here.


From this I deduce that having sufficient cash to withdraw means having sufficient cash above the original target cash allocation. In the example we started with 10% £10,000 cash allocation, so in theory there was sufficient to withdraw without selling anything. But the principle of the rules means you sell anyway and allow cash to become overweight. This is interesting as while a cash buffer is reassuring too much cash normally becomes a drag on returns.

I am planning to implement the Guyton Klinger rules but to get it right looks like it will require a spreadsheet to enter all the key values and run the calculations for how much to adjust the withdrawal amount and what to sell of which asset class and in which order. But it is proving a real challenge.

jamesd

The original analysis didn't use cash but yes, it would be cash above starting cash and cash would be ahead of bonds in what to "sell" to provide the income if overweight selling didn't do the job. Similar logic as equities over bonds, the cash is less volatile.

mrstout

Thanks again.

Is it OK that cash can get very over-weight versus target? Take this quite possible scenario: heavy falls -38% in equities and 10% gains on bonds.

Starting allocations:
Equities (65%) £65,000
Bonds (25%) £25,000
Cash (10%) £10,000
Total £100,000

End of Year
Equities -38% (£40,000) = 52% of £77,500
Bonds +10% (£27,500) = 35% of £77,500
Cash 0% (£10,000) = 13% of £77,500
Total £77,500

Sell any Equities with positive returns. None
Sell any Bonds with positive returns. Yes, sell all overweight allocation (next step)
Sell overweight bonds (10.5% overweight) = £8,125
Required withdrawal is £4,000, so withdraw £4,000 of £8,125 bonds sale and keep £4,125 in cash 'future withdrawals' reserve

Portfolio after selling:
Equities £40,000 (none sold) 52%
Bonds £19,375 (£8,125 sold) 25%
Cash £18,125 (£10,000 + £8,125) 23%
Total £77,500

Portfolio after withdrawing £4,000 cash for income:
Equities £40,000 (none sold) 54%
Bonds £19,375 (£8,125 sold) 26%
Cash £14,125 (£18,125 - £4,000 withdrawn) 19%
Total £73,500

Target Allocation: Equities 50% Bonds 40% Cash 10%
Closing Allocation: Equities 54% Bonds 26% Cash 19%

So at this point we are holding 19% cash - almost double the target allocation we started with.

Also, the cash balance is £14,125, after withdrawing this years £4,000, which means we are now holding at least enough cash to cover next years withdrawal of £4,000 (ignoring inflation/performance adjustments). So next year, what do we do? No selling of any assets would be necessary, so do we just rebalance back to the target Equities 50% Bonds 40% Cash 10%?

jamesd

Just continue to follow the rules. In many big equity downward move situations a year later equities would have done a lot of recovering.

If you are very confident that it is just a typical short term drop you might choose not to follow the rules and try to buy low on equities. But since there's nothing to prevent further drops this is an increase in risk level. Or you might have determined in advance that your initial cash allocation would be used for this purpose and move some of it into equities.

mrstout

jamesd wrote: »

Just continue to follow the rules. In many big equity downward move situations a year later equities would have done a lot of recovering.

If you are very confident that it is just a typical short term drop you might choose not to follow the rules and try to buy low on equities. But since there's nothing to prevent further drops this is an increase in risk level. Or you might have determined in advance that your initial cash allocation would be used for this purpose and move some of it into equities.

When you say 'just continue to follow the rules' does this mean the rules (in fact the whole approach) of Guyton Klinger does not require any rebalancing of assets? I cannot find any guidance on rebalancing in their papers or other commentary. Not rebalancing any more would take some getting used to, and a great deal of trust that the portfolio sustainable withdrawal rate performance achieved and tested on monte carlo simulations assumed no rebalancing.

"If I am confident it is a short term drop..." I don't want to try to time the market. Even in drawdown if I rebalance it would be for the longer term benefit I have seen time and again during the accumulation stage and not about timing the market. If it drops further, then rebalance again next year and keep doing so - if the Guyton Klinger rules allowed for rebalancing.

My instinct is to rebalance if there is a heavy overweight in cash. If I start with a target allocation of 10% cash this would represent over 2 years potential withdrawals at 4% p.a so I don't think I would want cash to exceed 15% of my portfolio. Maybe I have to start with a lower base of cash, 5% and increase bonds to 30%. My concern is that if all the positive outcomes using monte carlo simulations for Guyton Klinger are based on following their rules, and their rules never rebalance, then rebalancing will change the outcomes. But for better or worse? The Guyton Klinger approach appears to have a lot going for it but this is a dilemma when it comes to following it.

westv

mrstout wrote: »

My instinct is to rebalance if there is a heavy overweight in cash. If I start with a target allocation of 10% cash this would represent over 2 years potential withdrawals at 4% p.a so I don't think I would want cash to exceed 15% of my portfolio.

10% is only an example not a set in stone percentage. Also one benefit of following the rules is that the withdrawal rate can exceed 4%.