Continuing from Part 1, Part 2 and Part 3 in this series we now actually come to what was named the "Preet Principle". The name actually doesn’t make too much sense, it was more because my colleague wanted to come up with a name to describe this strategy similar to the "Smith Manoeuvre" or the "Singleton Shuffle". There really isn’t any "principle" – except perhaps taking advantage of piggy-backing refunds… which will become clear as you keep reading.

Nothing in the strategy is uniquely mine. I simply took existing strategies and put them all together, created a more formalized step-by-step process, and then ran it all through a probability calculator to quantify the risk-return of the strategy.

**The Background**

The strategy is not for everyone, which I think I’ve made clear in the previous parts to this series. It was originally developed for high tax-bracket investors as a means to both accelerate the growth of their net-worth through leverage and through tax reduction. Further, it is aimed at higher risk tolerant (aggressive to speculative) investors with lots of free cash flow.

I’ll begin by introducing a base case study for the traditional saver, compare that to an alternative leverage strategy, and then introduce the "Preet Principle" – which again, is not actually a principle, but rather a strategy.

**Assumptions**

Out test investor is currently 40 years old and will be retiring at age 65. He will live until 90 years of age, with full entitlement to CPP and OAS. He lives in Ontario and earns $250,000/year which puts him into the top tax bracket of 46.41%. He has absolutely no retirement savings as of today, but has committed $1,000/month going forward to his retirement plan. He is an experienced investor, but has spent all his non-registered investments on toys and the like.

Any refunds he receives from RRSP contributions or interest deductibility from investment loans is ours to use as we see fit.

His investment portfolio has a long term rate of return of 8% with a standard deviation of 12.5%. For non-registered purposes, he will use corporate class investments and as such his returns are calculated as 0.5% interest, 0.5% dividends, 0.5% realized capital gains distributions and 6.5% deferred growth. Any loans are calculated with an interest rate of 7%.

For calculating the impact of varying return patterns of his investments, we will run 500 different return patterns for his portfolios for each scenario assuming that a trial is successful if he does not have to tap into a line of credit for more than $15,000 in any given year. If he has to do this in any one year of his life – the trial will be considered unsuccessful. We will also randomize the life expectancies so that the computer will simulate living both longer and shorter than 90 years.

**Base Plan – The Traditional RRSP Saver**

The base plan will serve as a reference point for the two strategies that follow. In this base plan, our test investor will save his $1,000/month to his RRSP account. That adds up to $12,000/year going to his RRSP which will generate a tax refund of $5,569. This $5,569 will then be deposited annually to a non-registered account starting the year following his first RRSP contribution (since the refund won’t be owing until then).

So to sum up the base plan:

1. $1,000/month to his RRSP.

2. RRSP contributions generate a tax refund of $5,569.

3. Tax refund contributed to a non-registered portfolio.

Based on this, our test investor would have a healthy after-tax annual income in retirement of $49,875/year in today’s dollars, and if we run this through a Monte Carlo Sensitivity Analysis, we find that this plan has a probability of success of 72%.

**Comparison Strategy – A Straight Interest-Only Leverage**

In this case, our test investor will take out an interest-only loan for $171,428.57 to deposit to his non-registered account, since a loan for $171,428.57 would cost $1,000/month in interest payments. He would generate a similar refund as with the RRSP contributions since the interest on the investment loan is tax deductible as well. His refund would also be for $5,569. BUT in this case we are going to put this into his RRSP on an annual basis from this point on, which will generate ANOTHER tax refund. His $5,569 contribution will generate a further tax refund of $2,584 per year. This $2,584 is deposited into his non-registered investment portfolio. (You might want to have two non-registered accounts for this strategy – if you ever need to make a withdrawal, you would take it from the account where you deposited the annual refunds so you don’t affect the tax deductibility of the investment loan.) In order to take advantage of the refunds right away, you would actually set up an annual RRSP loan of $5,569 so that you could get your expected refund in one year earlier from the get go. This will be more clear when we list the steps of this strategy:

1. Take out an investment loan of $171,428.57 – costs you $1,000/month in interest (or $12,000/year).

2. Tax refund on interest payments equals $5,569.

3. Take RRSP loan out for $5,569 before RRSP deadline.

4. RRSP loan will generate another $2,584 tax refund.

5. Total tax refund for first year will equal $8,153.

6. $5,569 will be used to pay off RRSP Loan, the balance of $2,584 will be contributed to your non-registered account.

The beauty of this strategy is that you are using piggy-backing refunds. In this case, with an out of pocket cost of $12,000/year, our test investor is able to generate $8,153 per year in tax refunds. In this case, if our test investor only ever put the refunds into high-interest savings account that averaged 3.5% interest between now and the time he turns 65, he would have $328,672 – more than enough to pay off his interest-only loan, PLUS he would have whatever the original investment of $171,428 grew into over 25 years. So again, to re-iterate, the piggy-backing of the tax refunds help offset the risk of the leverage, and if you assume a long term rate of return higher than 3.5% on the refunds it only gets better. In fact, that is what we have done by assuming an 8% rate of return – let’s look at the new results:

Based on this strategy, our test investor would have an even healthier after-tax annual income in retirement of $62,700/year in today’s dollars, BUT if we run this through a Monte Carlo Sensitivity Analysis we find that this plan has a probability of success of only 65%.

One main point to remember, we have increased his annual income in retirement by a MASSIVE $12,825/year AND his out of pocket cost for each strategy is still only $1,000/month. But the drop in probability of success has been reduced to 65% and that may not be an acceptable trade-off for some. What if we were to calculate the probability of success for this strategy drawing the same income of the base plan of $49,875? It IS slightly higher at 74% (versus 72% of the base plan). That’s more like it!

There is a big caveat however: As we know, we reviewed why most leverages fail in Part 3. To briefly sum it up, leverages usually fail because people tend to do the following:

1. Leverage at the top of the market.

2. Stretch one’s self too thin with the monthly payment.

3. Have a lack of discipline or investment knowledge/experience to keep going in the tough times.

Further, the interest rate risk can be divided into two arguments. If inter

est rates go up, there is more danger of stretching yourself to the point where you cannot support the payments. The other argument is that as interest rates go up, the differential between return rates and the interest rates may not make a leverage feasible – and that is true in most cases, but with the piggy-backing refunds, don’t forget that this risk is slightly mitigated by being able to get more of a refund as well since the amount of your write-offs go up as the cost of borrowing increases.

So with that all in mind, I will now, FINALLY!, present a possible alternative:

**The "Preet Principle"**

It can be broken down to two parts. One part is the screen for qualifying potential candidates for this strategy and it is:

1. Your TDS Ratio must still fall at or under 40%. TDS = Total Debt Servicing Ratio.

2. You would ideally have been through a full market cycle during which time you were actively investing.

3. You need to wait until a correction of at least 10% before engaging the strategy.

4. You are either an Aggressive or Speculative investor, with respect to your risk tolerance.

5. You are comfortably in the highest tax bracket, have a high net worth, and have other assets.

6. You have a minimum 15 year time frame.

7. You will own between 20-40% of your portfolio in fixed income investments.

8. You expect the market to have another correction of 10-20% in the next 5 years.

I will describe the second part of the strategy before commenting on the screen:

After you have met the screen requirements, you would then engage a hybrid of the straight leverage and the traditional savings strategies, such that instead of $1,000/month going to either strategy, only $500 would go towards supporting the interest-only leverage, and $500 would go to monthly contributions to your RRSP… FOR AROUND THE NEXT 5 YEARS. At 5 years (or a different time as you will gather from below), you would then engage the second half of the original leverage for $1,000/month in total interest costs. (So at no time is your monthly commitment anything but $1,000/month).

I say for the next 5 years, but it was an arbitrary number because really you are going to monitor the strategy for the next few years looking for "something" to happen. By splitting the strategy up you are accomplishing a number of things, most of all: creating flexibility to react to bad events or take advantage of others.

1. Market has a big correction – in which case you would double down and convert the $500/month RRSP contribution to a second interest-only loan bringing you up to the original $1,000/month interest payment. This is in effect taking advantage of the market being "on sale" by making a lump sum purchase at this time. It also is sub-consciously convincing you to hope the market has a downturn so that you can take further advantage of it – which can help you to sleep at night – I know it sounds weird, but I LOVE when markets are going down – the opportunities that present themselves can be quite good.

2. Changes in interest rates – if interest rates go up, then you are not struggling since you can just reduce your monthly contributions to your RRSP to afford the increased interest payments.

3. Emergencies – if you have some sort of cash-crunch, you can just stop your $500/month contributions at anytime without explaining nuthin’ to nobody!

4. Market performs really WELL – then you just collapse your leverage to lock in your gain. You would switch back to $1,000/month going to your RRSP until the market has a correction. At that time you would again engage the 1/2 and 1/2 strategy, waiting for a second correction to trigger the full $1,000/month leverage.

So, let’s put it into perspective by running the numbers. I will assume our test investor uses the 1/2 and 1/2 method for 5 years and then converts to a straight leverage at year 5. (Note that the amount of tax refunds is exactly the same as the total amount of money eligible for deductions is always $1,000/month).

We already know that the annual income amount will lie somewhere between the base plan and the leverage based on what we know about leverage and return magnification… or do we?

Based on this strategy, our test investor would have an after-tax annual income in retirement of $58,450/year in today’s dollars (still a generous $8,575 more than the base plan!), AND if we run this through a Monte Carlo Sensitivity Analysis we find that this plan has a probability of success of 71%. If we further calculate the probability of success for the base plan income level of $49,875/year we get a massive **82%!**

**Let’s recap all three scenarios:**

Base Plan: 72% success rate at $49,875.

Straight Leverage: 65% success rate at $62,700; 74% success rate at $49,875.

Preet Principle: 71% success rate at $58,450; 82% success rate at $49,875.

**BUT, I will tell you that the success rates for the Preet Principle would in reality be even higher still.** This is because what the computer does not know is that we are waiting for a correction to engage the strategy – this increases our chances of success going forward. It also does not know that we have the flexibility to trigger the second leverage on a further correction, or collapse the initial leverage on a strong bull run whenever we want. Because of this additional information, I would speculate the success rates are quite a bit higher for the Preet Principle than stated above, but I just don’t have the foggiest idea how to model that precisely.

Lastly, while the Preet Principle annual income amount is sacrificing some potential reward in the end (in the computer’s mind) versus the straight leverage, keep in mind that by waiting until a correction before engaging the strategy will serve to not only increase the success rate, but the annual income amount as well. :)

So there you have it, this is my take on leveraging to invest – which, remember, is potentially only suitable, potentially, for certain investors… potentially. :P

If anyone has additional insight or criticisms, by all means fire away and maybe we can make this analysis even more accurate or correct any flaws, etc. So don’t be shy with comments or questions good or bad. The more I have to defend a strategy, the better I learn how to use it! (Or better yet, know if it’s complete crap after all!) :)

Thanks for reading the series everyone. Have a great day and happy investing.

If you found this article of interest, please consider subscribing to my RSS Feed. If you want to learn more about what an RSS Feed is, click here.

For special deals for readers of WhereDoesAllMyMoneyGo.com (that’s you!), please visit the "Deals For Readers" section.

Great series!

Mike

Thanks for the kind words Mike!

Preet

Hi, I’m wondering about some of the screening requirements, such as highest tax bracket. Would you mind to explain? Thanks!

Hi Cindy, no problem.

The higher tax bracket will generate larger tax refunds, which help mitigate the risk. So in my example where $12,000 out of pocket was creating around $8,000 in refunds which were re-invested in one form or the other, the effect is less pronounced with lower tax bracket individuals.

For example if someone were in a 22% tax bracket, the $12,000 interest write off would generate $2,640 in refunds, which would in turn generate another $580.80 in refunds for a total of $3,220.80.

Also, if you are in a lower tax bracket, you face the possibility that you will write your income down to a lower tax bracket – and the refunds will be less.

This won’t happen when you are "comfortably" in the top tax bracket – and also made for an easier analysis.

I have a spreadsheet I whipped together one day that allows you to play with interest rates, return rates and marginal tax rates – I’ll post it up and you can play with it.

You will find some interesting results when you play with the numbers… stay tuned! :)

I thought that the contribution to the non-leveraged non RRSP portfolio was kind of weak in all cases. It is relatively easy to show that the RRSP dominates the returns of Non-RRSP Non-Leveraged portfolios on an after tax basis in almost all cases so you might as well remove these from the analysis. There are two ways to remove the NREG component, either decrease your withholdings to the CRA (if you are self employed this is easy just pay slightly less in installment payments, if you are employed you can write a letter to CRA requesting they allow your employer to withhold less from you), or you can take out an RRSP loan. Basically you want to have spent your entire refund before you actually get it.

Instead of paying $12,000 in interest and using an RRSP loan of 5,569 you instead take a loan of $10,392. then your total refund is (12,000 + 10,392)*.4641 = 10392 which should be spent on paying back the RRSP loan. (This figure could be altered to take into account interest and fees on the loan).

Anyways the addition of this strategy will help out all of the strategies listed therefore it will not change the conclusion of the argument, it will just allow the person to either take out more cash each year without increasing the probability of failure or it will allow you to take out the same amount of cash with a reduced probability of failure.

Hi CommanderT – I’m not sure I understand your first paragraph. Are you saying that there should not be any non-registered savings in all scenarios because the after-tax return for the RRSP will be better?

If so, this is not necessarily the case once you factor in CPP, OAS, pension income, etc. as those are all sources of taxable income at your marginal rate whereas you are only taxed on the growth of your non-registered assets.

It really depends on a number of factors (what your projected income will be, your current and future marginal tax rates, timeframe, expected rate of return, etc.)

I would agree that most of the time the RRSP will be a superior structure though – but not always.

With respect to using an RRSP loan for the amount you suggested, I would be most happy to run the projection to show the difference, but I didn’t want to run off in too many directions all in one series. I did address this strategy in some other posts on the blog though – so the regular readers will be up to speed on that (well, at least MY view on it).

Also, the main reason there is a non-registered contribution was that the strategy was initially developed for a doctor who was maxing their RRSP but had more cash to invest so the original strategy was for a investment loan with interest payments large enough that the refund would max out his RRSP every year – therefore we would have no more room to use in the RRSP and would HAVE to use the non-registered account.

If you have a specific set of numbers you would like me to use (and indicate what scenarios you would like me to run), I will post the results as a comment in this thread.

Thanks for reading and commenting…

Preet

I was under the impression that withholding tax can be reduced automatically by the HR department only if the RRSP contributions are made directly to the company’s sponsered registered plan (and they deduct it directly from your pay). That is why the government has the T1213 form.

If you are making RRSP contributions early in a lump sum (or through a pre-authorized purchase plan) and it is not through an employer plan, you can fill out the T1213 form (available at the cra website) and send it in to the nearest tax service office. They will review your request. If approved they will send you a letter of confirmation which you then forward to your HR department. They will then reduce your withholding tax for the requested tax year.

Hi Jon, you are correct. You can fill out the TD1 form (I think) when you start a new job to reduce taxes at source, but once you are on payroll, you will need to fill out a form and send to a tax office to have your payroll taxes adjusted as you mention.

Thanks for pointing that out.

Preet

Hi,

I guess that we have to consider all strategies in conjunction with the mortgage payments, which most likely our investor (and majority of us) has to make. In this case we can create a real projection of possible retirement life stile. Taking in account that the "Smith Manoeuvre" can be utilized to generate an additional cash to invest (using tax refund from RRSP + investment loan interest), a huge mortgage interest saving during this 15 years might significantly influent the final conclusions.

Hi,

One more comment… The last strategy assumes that an investor can take a leveraged investment loan or collapse leverage at any time. In reality, based on my experience, if a financial institution granted you an UNSECURED investment loan (I guess that majority of the leveraged investments are based on unsecured loans), they will force an investor to hold securities that THEY suggested for 7-10 years. I guess they think that a person, who takes $250,000 loan has no idea what he or she is doing and will sell everything with losses in a panic mode when the market condition gets ugly… Of cause, an additional loan to take an advantage of good buying opportunities is out of questions…Thus, it looks like to me that the second scenario – Comparison strategy has a better chance to be implemented and managed.

P.S. Thank you for fantastic series…

Thanks for the insightful comments Acorn – you are right in that many lenders might put some restrictions on what exactly you can invest in if you go to them for an investment loan. One way around that is through using a line of credit, then you won’t have that specific hurdle.

With respect to your first comment – again you are right in that there are many options available to look at and it is best to really run through favourable and unfavourable scenarios of each, compare them, etc. I’m sure someone out there has a combination of this and the SM going at the same time! :)

Thanks for reading and commenting.

Preet

Hi there,

This is a very interesting series. The tax refund from the RRSP could go into a TFSA instead of going to a non-registered acct. Of course when the series was written we didn’t have TFSAs, what to do you think of that alternative?

Also, what do you think of getting a HELOC instead of a no-margin interest only loan?

One more thing, if you have a non-earning spouse and you contribute to a Spousal RRSP, after 3 years your spouse can withdraw some money or all of it depending how much it is so it doesn’t get taxed. Afterwards though, could your spouse give you that money as a gift so you can reinvest in the spousal RRSP and get another tax refund?

Thanks in advance!

Eugene

Hi Preet,

I realize this is quite an old post and you may not see comments posted here (as you will be even busier in the future – congratulations on the show!). I was wondering about the Monte Carlo risk analysis you used for your calculations. It seems it is based on the stock market following a normal distribution. I suppose my question is, does a normal distribution reflect the stock market returns closely enough for risk calculations to be useful?

Thanks if you (or anyone masochistic enough to be reading through all your archives) can answer as I have been wondering for a while.

@Charlotte S Monte Carlo does use a normal distribution, and that does not reflect the realities of the stock markets. Is it close enough? I’m starting to believe the utility is less for lump sum investments but good enough for investors in the accumulation phase.