Growth Rate in Drawdown

shinytop

Hopefully a simple question. I’m playing around with drawdown models and trying to come up with a real (post inflation) growth rate. The pot will be there (hopefully) for 20+ years. If I were to decide that 20% cash (in ISAs) and 80% in a SIPP in a 60/40 equity/bond fund (e.g. HSBC Global Balanced) was the right level of risk for me, what kind of growth could I assume? I have modelled it in Trustnet and can compare with CPI+1%, 2%, etc., but can only go back 5 years. Would CPI+2% be reasonable for the long term? This is only for an indication to see how much might be left when I pop off, not to calculate how much I can safely withdraw. Most of my income will be from DB+SP. I won’t need large cash sums for anything specific and I’d only sell equities/bonds to replenish cash.

Linton

In my retirement planning models 15 years ago I assumed inflation at 3% and investment return at 4%, ie 1% real return. This proved to be extremely pessimistic on both counts. Which is good, one doesnt want to get anywhere near over-optimism.


In practice it does not matter much as long as you start off with moderately pessimistic assumptions and regularly update your model with reality. You can soon detect if things are going off course well before it becomes a real problem and take corrective action.


So yes, in my view CPI+2% is a reasonably cautious starting point, but you should carry out a sensitivity analysis to understand the bounds at which your plan would fail.

bostonerimus

shinytop wrote: »

Would CPI+2% be reasonable for the long term?

That's sensible, but test your plan with la lower return and see how you do. Try to come up with ways to adjust the plan for the lower numbers to make it work.

Thrugelmir

Challenge you face to achieve your goal is the current level of interest on offer on cash deposits, and likewise on the returns available on bonds.

With an allocation of 48% of your portfolio to equities. You'll need a consistantly good return over a 20 year period to do the heavy lifting.

Historical data needs to be considered carefully. As the past decade or so has benefited from the lax fiscal policies of the Central Banks. Simply to keep the global financial system afloat. Over time the reversal is going to happen. With the Fed being the first last year, though already has put it's plans on hold. Being in unchartered waters no one knows what the outcome is going to be.

cobson

The rate of return that you use will depend upon the type of modelling that you are doing, e.g. you would need to use a higher rate of return with a Monte Carlo model than with a spreadsheet, as the volatility built into the former would introduce drag, and you will want a high percentage of success. This is illustrated in the last chart on here:

http://www.theretirementcafe.com/2014/07/half-right.html

A 5% rate of return with 11% volatility and a 90% confidence rate equates to a 1.9% rate of return in a simple spreadsheet model. Relying just on spreadsheet models, even assuming a low ‘safe’ rate of return, can be dangerous as it hides sequence of return risk. More here:

http://www.theretirementcafe.com/2014/08/spreadsheets-and-sor-risk_25.html

[Deleted User]

Even if you guess correctly (a big if), what really matters is the sequence of returns. Early stock market losses can completely undermine portfolio if the withdrawals are at a constant rate, even when long term returns are good. Conversely, you could be leaving way too much money and underspending, if the early returns are good.

Variable withdrawal strategies mitigate the risk of failure and increase funds available to spend. And you don’t need to try and guess future returns.

shinytop

Interesting replies, thanks. As I said, I'm in the fortunate position that I don't need to withdraw a set amount or need amounts that would deplete the pot to zero. As I understand the article, the point that 2% might not be enough is relevant where the retiree is pushing the SWR to the maximum (4 or 4.5%); I'm not.

In practice it does not matter much as long as you start off with moderately pessimistic assumptions and regularly update your model with reality. You can soon detect if things are going off course well before it becomes a real problem and take corrective action.

Yup, that's what I'll be doing. 2% works for me; in fact my original model used 1%. I have scope to be flexible and won't ever need to sell a lot of equities if I don't want to.

pensionpawn

Linton wrote: »

In my retirement planning models 15 years ago I assumed inflation at 3% and investment return at 4%, ie 1% real return. This proved to be extremely pessimistic on both counts. Which is good, one doesnt want to get anywhere near over-optimism.


In practice it does not matter much as long as you start off with moderately pessimistic assumptions and regularly update your model with reality. You can soon detect if things are going off course well before it becomes a real problem and take corrective action.


So yes, in my view CPI+2% is a reasonably cautious starting point, but you should carry out a sensitivity analysis to understand the bounds at which your plan would fail.

Are you willing to share your actual observed growth and inflation rates, it would be interesting to know? I ask because, and I acknowledge that there are many on this forum who have greater experience and have done more extensive analysis than I have, when you can achieve 2.6% via an online 5 year bond some of the rates being suggested for modelling growth seem very conservative. Personally since I opened my first pension 30 years ago my observed growth rate is ~7% (without deposits) across all my longer term funds. I appreciate that long term averages contains significant troughs and peaks and the danger of a trough occurring shortly after the commencement of retirement will be damaging and needs to be planned for. That said one of my guiding principles is to find the right balance between accumulating the largest possible pot of money and accumulating the required pot of money to support the desired retirement lifestyle.

Linton

pensionpawn wrote: »

Are you willing to share your actual observed growth and inflation rates, it would be interesting to know?

Since I retired nearly 14 years ago the average IRR ( which accounts for buys and sells) is 6.4% My portfolio consists of a wide range of different types of investment and so the rate of return is less than a pure equity one would achieve. Note that this time period includes the Great Crash.


As to inflation, I have found CPI more than adequate to match my actual expenditure. CPI has averaged 2.3%. So both my assumptions of 3% inflation and 4% return have proved to be very pessimistic. However I have not changed them for future planning. Who knows what the futrure will bring.

Linton

Deleted_User wrote: »

Even if you guess correctly (a big if), what really matters is the sequence of returns. Early stock market losses can completely undermine portfolio if the withdrawals are at a constant rate, even when long term returns are good. Conversely, you could be leaving way too much money and underspending, if the early returns are good.

Variable withdrawal strategies mitigate the risk of failure and increase funds available to spend. And you don’t need to try and guess future returns.

If you are planning you do need to guess future returns since you need to have some degree of confidence that they will fund your desired lifestyle.


Personally as a retiree I have zero interest in variable withdrawal strategies. Withdrawal in practice is a tactical decision based on circumstances at the time and the results of simple long term planning. In my view it is a mistake to confuse the assumptions you need to make to achieve a useful plan, and a variable withdrawal strategy is such an assumption, and what you actually do.

[Deleted User]

Variable withdrawal strategy assigns a percentage you can withdraw - of your total fund in any given year. As long as you follow the schedule, you succeed. If the portfolio halves next year, your “income” halves too. That’s always a possibility. In fact, most people cut their spending in such circumstances.

Retirees who don’t use this approach tend to underspend, live the lifestyle below what they can actually afford and end up with too much money they can’t spend at the end. And it make sense - assuming a constant rate of withdrawal upfront leads to failure in too many scenarios. So they have to be too conservative.