[steve41]

People have to be aware, and I expect this would be explained in Jim's book, that these quick rules of thumb have to be taken with a grain of salt. The "asset to withdrawal ratio" (the "4% withdrawal rule" by another name) doesn't consider real life situations. If everyone was 65, of normal health, and lived in a vacuum (no expectation of entitlement income, no outstanding loans, no part time retirement income, no desire to pass on an estate, no expectation of a future capital gain such as selling the cottage, no anticipation of a future cash call or special expenditure), then these rules of thumb might have a use.

The fact is, our life is not that simple... our capital is taxed in very different ways, and age is very important. For instance, our "A2W ratio" has to be much lower prior to age 60 and age 65 to accommodate CPP&OAS bridging, and of course, as we approach 85 and beyond, the A2W approaches the number 1 (one) since in our final year we (hopefully) withdraw the last and final dollars from our retirement next egg.

The only number worth knowing is your burger quotient (BQ).... that lifestyle (after tax/after inflation) which if followed, will (just) see you out to a certain age. It requires some computation because it includes the effect of income tax, its effect on the different forms of capital (reg/nonreg/equity/tfsa), other discontinuous financial entities such as loans, future lump sum cash infusions, pensions, entitlements, etc... however, this is a calculation that everyone needs to do.

It is your life and future well-being as well as your heir's... this is more important than applying a simplistic "4% withdrawal" or "asset to withdrawal ratio".

BTW... unless you are way up in the HNW stratosphere, if you expect your investment adviser to volunteer to do this, think again. This is an exercise you should be doing on your own.

A financial plan starts with your Visa bill(groceries/gas/etc) and works back from there. The industry wants you to be afraid of uncertainty... they would prefer that you stick to simplistic adages such "max your RRSP pre-retirement", 4% withdrawal rate post retirement, rather than getting involved with time-consuming analysis such as described above, which earns them no fee.

Cynically,

Steve

 

[steve41]

FWR (financial webring forum) Canadian based financial forum

Gummy's website Lots of tools/info/spreadsheets... definitely a 'must see'.

SWR -sustainable withdrawal rate

 

[steve41]

Does anybody have any other alternative approaches/guidelines to arriving at a SWR?

Read my post upthread.

The problem with safe withdrawal rates is they apply to individuals living in a vacuum. I could take two people with exactly the same situation... same savings (amt and type), same age, same entitlement situation and same size of loan. They could have completely different SWRs applicable.

How? Simple.... one has a loan with a 3 year amortization and the other with a 10 year. These 2 different loan payment levels put each individual's SWR strategy into different territories. (depending on the scale of the loan amounts they could be quite different)

The best strategy, IMHO is the After Tax Income (ATI) Monte Carlo. This differs from the normal MC in that it uses a fixed term 'die-broke' model rather than a fixed withdrawal rate model. The latter takes a withdrawal level and runs it out over a randomly varied rate regime, measuring the years at which the funds run out. The results tabulate the probability that your funds run out at various ages. The fixed term (ATI) model, rather than measuring years to run-out, measures net income... that after tax income (ATI) which, if sustained, will exactly run out at a particular age (95 or 100 say).

The reason I find this more sensible is that it relates exactly to lifestyle and the budgeting process, and it includes the effect of all other non-investment entities... loans, CPP, income tax, ... i.e it is much more inclusive that the simple 'level-withdrawal/investments-only' MC model.

It is far more meaningful to express... "your sustainable lifestyle, probabilistically, varies between a high of X, low of Y and a median value of Z" This relates to something I can control (budget for) rather than the kind of meaningless... "your funds may run out as early as age 69 or as late as 101" How does that relate to how my life (consumption of beer, groceries and gas) will unfold?

It takes a quantum increase in computing power (you couldn't do it with a spreadsheet), however with the speed of the present day PC, it is quite do-able.

You asked.

 

[steve41]

I should have clarified that the 'die-broke' paradigm doesn't have to mean 'run out of capital'. It could just as easily have been... "pass on a prescribed estate (1/2 Million after tax say) and then have the capital just exhaust"

I don't know about you, but if went into an adviser and he said.... "forget about the uncertainty regarding rates, taxes, details such as your loans, CPP status, a future capital gain... let's just get into picking an investment opportunity."..., I would not be impressed.

Many advisers look at inclusive cash flow financial planning as a nuisance. Too much data gathering. They (and the client in some cases) are just lazy... dredging up info on pension and CPP expectations, real estate details, loans, etc.... is a drag. It's much easier to get down to the nitty gritty... "I've got a handle on a new 'can't miss' MF. How much can you afford to plunk in your RRSP this year?" (I am sure this doesn't happen all that much, but you get my drift)

 

[steve41]

So how do I figure out what my (retirement) number is?

Simple. To determine your "number", you enter your assets... current savings (rsp/tfsa/nonreg) as well as your "career asset" (gross pay/pension and retirement age), future assets such as selling the cottage, downsizing your home, liabilities (loans), a 'how lucky do you feel?' (horizon age), and the amount you want your estate to net if you make it out that far (it could be zero).

Finally, pick one or more rate trajectories (hi/lo/average), sit back and wait.

The result will show how much you should be contributing (it may not be the RRSP or TFSA max BTW), how much you should be drawing down after retirement, and of course, your "number"... what lifestyle (after tax, after inflation) your plan will deliver year over year.

The only other "number"... the size of your nest egg at retirement... will get spit out as well, however I don't find that "number" particularly useful.

 

[steve41]

OK... let me uncomplicate things. (I wrote the program, but sometimes it confuses even me)

The program is essentially used as a one shot deterministic projection. You specify all the aforementioned... salary, RRSP, loans, horizon age (95 say) and a rate estimate and the program proceeds to find a constant after tax income (BQ) which will exactly run the capital out at that horizon age (or exactly deliver a prescribed net to estate).

It returns a single deterministic projection, a cash flow... monies flowing in, out and between the two main capital pools... (reg and non-reg) which forces an exact level ATI/BQ while running the capital out at that age 95.

That is the program as most users run it. Now for the montecarlo bit...

The rate estimate, rather than a single number, can be, in fact, a continuum... a column of annual rates, which most users will keep constant over time. However, there is nothing to prevent that rate column to be randomly varied for each year.

This is what the monte carlo option does. It runs that single deterministic model multiple times, using a different randomly generated rate column, coming up with a different ATI/BQ each time.

Each time it runs, it keeps track of those ATI/BQ stats and displays them probabilistically (hi/lo/avg/med and a frequency distribution).

The montecarlo option takes a minute or more to run, because it is essentially running the program 100 or so times. What would take 2-3 seconds (the deterministic model) now takes a minute or two, and the only output is the ATI/BQ distribution. A normal deterministic run, which is what most users do, details all the cash flows, while the montecarlo run simply tabulates the ATI/BQ stats as a probability distribution.

Hope this makes sense.

To finalize the explanation, when I say a 'smooth constant ATI', you can contour that by having the ATI reduce to some % (75% say) after a certain age. This acknowledges the fact that some individuals may have a reduced lifestyle need in retirement.

Whew!

 

[steve41]

There is another calculation methodology the program uses. You specify all the same stuff as before... salary/pension contour, loans, horizon age, estate goal, etc, except you enter one more element... the desired ATI/BQ. What the program then does is determine the rate which will satisfy all those constraints.

So you can either...

-set the ATI/BQ and rate vector and determine when the capital will run out.
-set the rate and horizon age and solve for the ATI/BQ or
-set the horizon and ATI/BQ level and determine the rate

 

[steve41]

Leslie when she says "Safe" means 0% probability of failure

I would handle that using the 'under the mattress' method... setting the entire rate vector to zero. A 51 yr old earning 90K, retiring at 65, with a 100K loan (6%, 10 year) and 200K in his "mattress RRSP" will see a $29,208 'die-broke-at-95' ATI. (Taxed in BC, 2% inflation, full OAS, CPP at 65)

At a 6% rate, the ATI/BQ is $45,257. Hmmm... I wonder, is the 16K extra beer and groceries worth the agro?

 

[steve41]

This is what I find so silly about the 'number'. I can run the same 65 yr-old as a 'die-broke at 95' subject with zero rate of growth and get a number of 20.4

I can run him again and receive the same ATI/BQ (about $24K) and come up with a 'number' of 8.3

The only difference between the two projections was that in one case he was paying down a 100K loan over 3 years, and in the other instance he was paying it down over 10 years.... exactly the same die-broke lifestyle (BQ) but vastly different 'numbers' (20.4 and 8.3)

Does anyone see what I am driving at?

10 year loan
3 year loan

 

[steve41]

OK I could have changed my guy from two different loan strategies to two different CPP timing strategies... taking at 60 and taking at 65. The point is... here is someone with exactly the same age, financial assets, etc... except his "A2W number" can vary all over the map. The above example showed A2W values in one case 8, and the other case 20.

My number (the ATI/BQ) varied hardly at all. That's my point.... what is the significance of the A2W?

And more importantly... can a person get dangerously mislead by not understanding it correctly?

 

[steve41]

If no one needs a plan, why did the term 'financial planning' or 'financial planner' get into the lexicon?

Surely, spending a modest amount of money on books or software to determine some general level of spending/budgeting/investing in order that we don't start sucking air at age 70 (running out too soon) or have our rotten kids inherit a gazillion dollars when we die, is not too much to ask... given the way we waste money on other frivolities.

 

[steve41]

Garbage.... My folks had a set of tables (they were very common from what my Mom told me) These tables had three sections...

-mortgage/annuity
-sinking fund
-future value

Using these 3 tables, you could, based on what you had saved already and what you planned to continue to save until retirement... what you could look forward to in retirement. They were arranged by rate and length of term.

They taught the use of these tables in high school home economics!

Well before the computer even existed.

 

[steve41]

What you are referring to is simply the three finance equations that EVERYONE need to learn, but 99% refuse to.

have brain washed them into thinking they 'need a plan'.

Which is it? no one needs a plan, or everyone needs to learn. Make up your mind.

Using a set of tables, which people did before the computer or calculator were invented, or using a fully inclusive computer model... there is a need out there for people to get a handle on where they are going, whether they are saving enough (or too much), and what kind of lifestyle will see them out to some age or enable them to ensure thay have an estate to pass on.

The plan is revisited every year depending on how their investments have fared, any new taxes or financial products that have been introduced, and any changes that may have cropped up as far as retirement goals, estimates of the markets, etc.

Things change, we adjust and we tweak our plan. If you are content to wing it, fine. Many don't.

 

[steve41]

Along with the concept of financial planning, we find that some individuals 'budget'. No, they don't sit down every time they are about to go to the store and make a purchase, but at various times they (presumably) determine a budget to see whether they can afford a major purchase... can I buy a new car every three years? what should I plan to spend on a monthly scale such that I will be able to buy that new car, and still make it out to some reasonable age without sucking air.

Budgeting means striking a balance between spending and saving. If you choose to blindly plow ahead without any foreknowledge, fine... maybe you are blessed with scads of money... I'm not.

Using a calculator or a set of tables might suit some, however the world is much more complicated when you include the effect of income tax, bridging, inflation, partial retirement,... unfortunately, "common sense" doesn't always cut it. The mere fact that you are sitting at a computer and participating in this forum says you have access to more computational power than your parents could even dream about. Why not harness it in some small measure?

 

[steve41]

I can remember many, many years ago... I think it was 1st or second year algebra. It was a module on finite series. We went through all the time value of money derivations to the point we could derive them on our own (PV, Sinking fund, FV, annuity...). It was rigorous, but once you understood it, it was pretty trivial.

At the end of the module, our prof showed us how to derive the formula for an indexed annuity. That was pretty hairy, but the punchline he delivered went along these lines...." All the math we have just slogged through is pretty much useless... none of these formulae relate to the real world. For instance... why should we derive a financial plan and assume that the interest rate is going to be constant? (as we get older, we grow more risk averse and our rate expectation goes down) or why should pmt levels stay constant? Might it not be better to design an annuity which has a higher pmt (adjusted for inflation) in the first 10 years to account for the fact that our lifestyle requirements may take a dip in our latter years? or surely these investment cash flows don't exist in a vacuum. You will need more income in the years prior to 60/65 before your CPP/OAS kick in, or before your loan is paid off... fixed payment (inflation adjusted) annuities don't allow for that. Or how about a planned-for sale of your cottage 10 years out?"

The time value of money algebra is fine for teaching someone compound interest... the general implications of saving for retirement and subsequently living off the proceeds, but as practical, real life tools, they just don't cut it.

Now for the coupe de grace... Income Tax. The only element which has meaning (to me anyway) is the amount of cash (after tax) available to purchase beer, groceries and gas (as well as the occasional car every 5 years say)

The above-mentioned math will not solve that problem for the principal reason that the income tax formula is not linear... it is a complex calculation involving discrete tax brackets (indexed to inflation) age credits, loan interest deductibility, dividend tax credits.... it is a nightmare.

Why is it a nightmare, given that many of us still do our T1 by hand, you ask? Simple... the tax formula was designed to work from the top down, whereas for the purpose of financial planning, we need to start at the bottom (net income) and drive the T1 backwards. The simple question... "how much should I draw from my RRSP such that I net (after tax) exactly $30,000?"... sounds easy, but it isn't. Now throw in the reality that tax on our RRSP withdrawal doesn't live in a vacuum... we are taxed on investment growth on capital outside of our RRSP, maybe even at the dividend rate. We are receiving additional taxable income from CPP, a pension, an annuity. This is absolutely impossible to solve with even the most convoluted set of time-value-of-money formulae.

Thankfully, there is a way to solve these kinds of problems... recursion math. The way you would solve that simple 'how much to draw from my RRSP to return $30K after tax?' question can be done by hand. You simply continually shovel RRSP withdrawals into a T1 program until you get close to the answer. It can be time consuming... trial and erroring 10 , 20, 100 times until the exact $30K drops out, but it is do-able.

Before the modern day computer arrived, this would have been the way to solve the 'reverse tax' problem, but now, the computer allows us to solve the 'needs-based' (after tax driven) tax accurate financial planning problem quickly and easily. The bad news, is that spreadsheets are not sufficiently fast or flexible, however using a procedural language (C++, Basic, Fortran, ...) will solve the problem and allow it to converge in a reasonable (several seconds) period of time.

Sorry to ramble on like this, but is is kind of my life's (well the last 15 years anyway) work. I get sort of passionate about it.

 

[steve41]

I think the sad truth is that Leslie ("the luddite") is in the definite majority. I can't remember the statistic, but I read somewhere that the percentage of Americans who had a written financial plan (numeric-cash flow) was just over single digits.

I get either really depressed or enthusiastic about the opportunity.

Time will tell.

 

[steve41]

I once tried to construct a book which could be of general use and was comprised of several hundred tables. I used my program to generate the tables. I soon discovered after observing many disparate financial plans, that hardly any of these tables would be of the slightest use. Everyone is in their own complex financial universe.... retired/not, pensions (various types), savings (reg/nonreg/equity), loans (of various complexions) expectation of a future capital gain (downsizing home, selling cottage, inheritance), different income targets (buying a new car every five years...), estate goals (die broke, specifying an estate level on death).... it goes on.

Trying to make even the most remotely useful set of tables was a complete non-starter. I am afraid that the only way (other than using the good old "common sense" dartboard) is a single inclusive program which allows you to specify all the above variables in one integrated number crunch. (rather than running separate disconnected spreadsheet illustrators)

The point I was trying to make about the A2W ratio was exactly the same as my experience... I determined that two identical situations (paying off a loan in 3 rather than 10 years) gave two completely disparate A2Ws... 8 and 20.

 

[steve41]

What I can't figure out is that years ago, before investments were taxed and there was no CPP nor OAS, Some individuals actually used compound interest tables to get a handle on how much they should be saving, and, if they saved at a particular rate, how much retirement income they might derive from those savings. People actually did that, believe it or not. Markets were uncertain then as now, and yet they still muddled through using future value, sinking fund and annuity tables.

Nowadays, it seems that people don't plan... they leave it to their "financial adviser". They seem to have relinquished any thought of doing
it themselves. Granted, things are a bit more complicated... those simple compound interest tables have become unusable... tax has become a much more complex issue with RRSP tax deductions, tax on investment growth of non-reg capital, the discontinuous effect of entitlement income, loan interest deductibility, etc... but, nothing has changed in the sense that rates of interest/market growth are as unpredictable as ever.

Why do we seem to be avoiding DIY financial planning... a simple schedule of savings and withdrawals based on an estimation of rates so as to ensure a smooth, sustainable lifestyle out to a certain age? said he rhetorically.

 

[steve41]

My diagnosis is the lack of consistency. When you submit data using an annuity table or function key on a financial calculator, or enter your T4/T5 numbers into a T1 program, you will get the same answer each time you submit the same data.

Not so with the financial planning projection. There is a plethora of web calculators, spreadsheets, etc, each of which cover some aspect of the financial planning problem. They are all different (different coding, different levels of accuracy, etc) and for that reason we are confused and disappointed with the process, so we delegate it to a planner.

The problem is that the planner hasn't got an accurate or complete handle on our financial information... how our salary/pension might unfold over time, loans, non-financial info such as a future plan to say, downsize our home or sell our business, capital we might have in other financial institutions we haven't told him about, or our desire to ensure an estate to pass on,... most of the data required to source a comprehensive financial plan is locked in our own knowledge base, not in our adviser's.

Consistency is knowing you can enter the same data set (including tax based data) into a single program/process and obtain exactly the same results each time. After all, the rules (compound interest, inflation, CPP/OAS/GIS, taxation, RRIF and LIF minimums and maximums...) are all stipulated and cast in stone, so why shouldn't there be a single consistent result?

Granted, there needs to be some general rules... such as to always max/invest in your RRSP first, draw down your non-rrsp capital first.... but once the main parameters (rates, inflation, etc), are set, then there is nothing to prevent the same consistency you would get from using say, a QuickTax, TaxWiz or CanTax to prepare your T1.

Just my two cents.

 

[steve41]

A financial plan is not cast in stone, it is a continual exercise. Retirement age, rate expectations, to sell the cottage or not, leave an estate or not..... these are revisited every time something significant (external or internal) happens. As for 200 questions, I don't know about that many variables, but surely a dozen or so would not be that onerous. This is, after all, your future... are you going to be drinking imported beer, no-name beer, or no beer at all. Granted, when you are just starting out, DIY planning doesn't seem that important... you are busy raising kids, building a career, buying a house. My observation is that the mid-to-late career individuals engage in this stuff mostly.

Financial planning... tailoring your lifestyle as it is influenced by non-investment elements... tax, loans, real estate, salary/pension, CPP/OAS... is best done by you alone, at your leisure. Investment planning, depending how much you follow/understand the markets and risk, may best be farmed out to an investment adviser.