Blessed by the Potato

There’s something of an art when it come to reading financial statements as an investor: a lot of things are open to interpretation in how you value them, and your own valuation may not necessarily agree with that of the accountants who prepared the statement.

A great example was brought up recently by SBrunner. She runs a very strange site, where she provides all kinds of data on a company, almost entirely free of commentary. Just spreadsheets and facts. Recently she highlighted one of my favourite companies (and largest holdings), Canexus. Though she doesn’t explicitly say that she disapproved of the change in Nexen’s ownership, she did prominently highlight the large drop in “book value” that happened at the time.

The thing is, the transaction had no effect on the true book value. It was all just an accounting shadow play.

Consider a case where I own 1/3 of some tangible thing, and my friend Netbug owns the other 2/3. I can choose to subdivide my ownership of the thing into 100 shares and then sell those shares on the stock market. Each share is ownership of 1/100 of 1/3 of the thing. If we put a value on this tangible thing, say $100,000, then each share is worth $333.33 in this example.

Now let’s suppose that Netbug wants to raise a bit of cash, but doesn’t want to sell his entire stake in our shared mysterious thing. He converts his ownership into shares that are identical to the ones I created: he creates 200 shares, so each is still worth 1/100 of 1/3 of the underlying thing (or if you prefer unsimplified fractions, 1/200 of 2/3). He could then sell those shares into the stock market (where they would not necessarily trade for the underlying value of $333.33/share).

This is a bit of a complicated transaction, and would greatly increase the float… but at the end of the day, the book value of each share didn’t change: the thing that we own is still the same, and each share still represents the same fractional ownership of this thing.

Well, that’s pretty much exactly what happened with the “buy out” in Canexus: Nexen converted it’s 2/3 ownership of the underlying business into shares, and dumped those shares on the market. Yet SBrunner reports a massive decrease in book value associated with the transaction. If one were to judge that and add some commentary, you might say that it represented massive dilution, or may have been a poor takeover, or something to that effect… if it indeed were a take-over and if book value had in fact changed.

So what happened? SBrunner, for the record, didn’t get it wrong: the book value according to the financial statements did drop by a massive amount. If you look to the balance sheet statements before and after the transaction, divide by the number of shares, you get the values she reports. She’s just faithfully tracking what the accountants are saying.

The problem was with the way book value was reported in the statements. For some reason, when the shares represented a minority ownership, the book value was calculated in a way that did not involve taking the net of the assets-liabilities and dividing by the ownership stake. In other words, the books of Canexus before Nexen spun off their stake did not simply represent 1/3 of the underlying operating business. There was this line item for the value of the ownership stake, and it got in there at the level it did by some accounting wizardry. From what I can tell, goodwill without calling it goodwill.

It was fine for the auditors, and probably made sense to the tax people, but as an investor, why would you have used that number as your figure for book value, instead of taking 1/3 of the value of the underlying business? If you had calculated book value yourself using that method (and the statements for the underlying operating company were reported in the same annual reports, just a few pages further in), then the transaction of releasing more shares from the previously non-trading part of the ownership didn’t affect your personal sense of how the company should be valued at all. If you relied on the way the statements were prepared, then suddenly there was this massive shift in value under your feet, simply because of how the accountants reported the worth of a minority stake versus the whole pie consolidated together.

So this is one real-world case where you may need to come up with your own interpretation of the financial statements, rather than relying on what the accountants have prepared according to GAAP/IFRS.

Be sure to look at the first two posts in this series introducing the retirement spreadsheet, and discussing the details of the inputs and how it works.

Here, I want to stress the importance of using this spreadsheet to look at a large number of potential scenarios. It’s good to have an idea of what you think would be the most likely outcome, and with a bit of conservatism, what would be the most likely base scenario (where things would be that good or better for you). But it’s also important to know how the factors will affect your ability to stay retired, in part so you know what to watch for as retirement unfolds. Keep updating the plan every year or so, and watch for any changes in the world that will require a course correction on your part.

When looking at this with my friend, I went through something like 25 different scenarios. I’ll outline a few for you below, but it’s important to do the exercise yourself: come up with your own range of reasonable values for all the various inputs, and understand how they’ll impact your retirement.

What might be an important factor can depend on circumstances. For instance, if a large part of your spending needs are met by a pension (plus CPP/OAS), then above-indexed personal inflation can have a massive impact on your ability to meet your expenses, while you may be ambivalent about market returns.

Let’s take the example in the default spreadsheet: someone who has $540k in assets, $45k in spending needs, and $43k in pre-tax pension income. They should be set up for a pretty secure retirement: only a small portion of the assets need to be used to pay for expenses — basically just the taxes and for the occasional big-ticket item every 5 years.

If their personal inflation rate matches the CPI rate that the pensions and government subsidy are indexed to, this person could expect to live to 92 on their assets even assuming conservative investment returns (-0.25% real return on fixed income, and 3.5% real return on stocks). Even after the money ran out, they’d only have to cut their spending by about $10k/year to live off the guaranteed pensions (it sounds like a lot, coming off $45k, but perhaps not so tough when you’re 92, and not the end of the world).

Even really poor investment returns don’t change that answer too much: lowering stock and fixed income returns by a full percent lowers the out-of-money age to 84; lowering stock returns right to 0% only brings it down to 80 — younger than many may expect to live, but zero nominal stock market returns for that long is, IMHO, needlessly pessimistic. So in that scenario you’re fairly insensitive to market returns.

Inflation, however, can be a killer. If your personal inflation rate is just 1% above that the pensions are linked to, your capital gets burned through at 80 instead of 92. At a personal inflation rate of 2% above the indexed rate someone in this situation may not want to consider taking retirement at the age 60 assumed in the scenario, since they’ll be facing shortfalls by age 75.

For someone without the pension, more personal assets might be needed to fund retirement. Let’s say that we had a similar base case: $45k/year spending needs, same CPP/OAS for the couple, but put off retirement until age 65, and an extra $100k of assets (all in the RRSP, for a total of $400k there). In this case, the person would make it to age 95 with their assets using the base assumptions (2.5% inflation for both inflation boxes, 6% nominal stock returns, 2.25% nominal fixed income). Personal inflation changes still have an effect: a 1% increase over CPI reduces the age of security to 87, and 2% to 82; but the effect is not as severe as when you were watching the buying power of your pension disappear under those inflation assumptions. And, if your investment returns increase by even half the increase in your personal inflation rate (e.g., by investing in companies based in the sectors that make the things you spend most on), then there’s hardly any pain at all.

If even more of your income is tied to the market, then so too will be your ability to retire.

I made a little document with a paragraph or two for each scenario, gave them names, and also discussed how likely they might be to occur — or perhaps more importantly, to work out that way or better. I focused mainly on the scenarios I thought were less likely: for instance, I think 2% real returns on stocks will likely be on the low side, but it’s more important to have a conservative plan for poor returns like that than it is to focus on more likely cases (say, 4+%/year) and be disappointed by reality. The next decade may very well bring extraordinary investment returns, following this previous decade of underwhelming ones. But hope and optimism are not a plan.

Since selling the house to fund retirement was not to be part of the plan for taking early retirement in my friends’ case, I didn’t explicitly build it into the tool. But if you’re close to retirement and using this tool, you could easily consider the case where you sell the house right away (either to downsize or rent): just adjust the annual spending budget accordingly, and increase the investments. Or, to consider the case where you sell it (or some other personal asset like a cottage) you can go to the calculations tab and change the line for the age where you sell it, and add in the extra assets. For instance, if the calculator says your retirement will be squeezed by age 78 under some scenarios, you may think that’s a perfectly fine outcome since the tool isn’t taking into account other assets that you have the option of selling at that point.

Also remember that financial planning isn’t a one-and-done exercise: it’s important to review your plan periodically to see how well you’re sticking to it, and how the world around you is unfolding relative to what you assumed. Then to make course corrections as necessary to keep moving towards your target.

Be sure to read the first post on the Retirement Calculator.

Here you’ll find some details on how I designed the retirement calculator. If you’re industrious, you’ll probably see some areas where it can be tweaked and improved. It is, in my opinion, a good and handy tool to have, but I know that the future is so uncertain that having a decent approximation is probably as good as you can expect to rely on, so there are definitely diminishing returns to trying to squeeze out more exact calculations.

The calculations basically assume that you’ll have some level of after-tax spending needs that you’ll be paying for out of a combination of pension, CPP, and OAS income as well as your savings/investments across non-registered, RRSP, and TFSA accounts. Those spending needs will increase year-by-year with inflation, and your investments will grow with stock and fixed income returns. Everything on the calculations page is run in real dollars, but if you’re just using the front user-interface sheet, you won’t notice those (the exception is taxes, which are paid on nominal returns).

A certain amount of cash will be kept on hand in non-registered fixed income to cover spending needs: at least enough to cover the following year’s spending needs. The RRSP is assumed to be in a mix of fixed income and equities, with that ratio being set as a multiple of spending needs (by default, 5 years of spending needs in fixed income) — you could adjust it to use one of the age-based asset allocation rules-of-thumb if you like. Regular withdrawals will be made from the RRSP into the non-registered accounts, whether they’re needed or not, with more being withdrawn if it’s needed for spending. From age 71 onward, it is a rule of the RRIF that a certain percentage must be withdrawn — the spreadsheet incorporates this, but also assumes smaller percentage withdrawals starting immediately upon retirement. The TFSA and non-registered equities are assumed to be in 100% equities, and will be used up before additional RRSP withdrawals are made.

The spreadsheet is fairly generalized: you can use it as a single person, or combine your expenses and income with your spouse, it shouldn’t matter either way.

In detail:

Spending budget: Simply, how much you’ll want to spend each year. If you don’t really know what you’re going to need to spend, then you can use a rule of thumb, such as a percentage of your spending pre-retirment (e.g.: 70%). I recommend starting with a full budget to see where the money is going, which would also help you give a better idea of how inflation might affect those spending requirements later.

Spending 5-Year Bonus Amount: Budgets are rarely smooth, in my experience. Plus when drawing up your budget for an average year, you may forget rare large expenses, such as buying a new car, or having a major house repair bill. This factor will build in a “bonus” spending amount every 5 years in the amount you specify. For example, if you put in $20,000, that might represent a new ~$20k car every 10 years, and a renovation/repair project every 10 years, with the two alternating every 5.

Pension Income: Few people are lucky enough to have a defined benefit pension, but if you managed to accrue a few years of service in one, you can put in your annual pension income here. Many pension plans include a top-up for those who retire at 60 to partially cover CPP that isn’t activated until age 65. If you want to fiddle with retirement age, you’ll likely need a copy of your pension statement or a pension calculator to figure out how to adjust the amounts for different ages. Any pension income is assumed to be inflation-indexed to the CPI rate.

CPP: Your annual Canada Pension Plan earnings: if you’re close to retirement and want to be precise, you can contact Service Canada to get a statement of what your CPP payout will be. They get put in at age 65, though it is possible to claim them early or defer them. Not hard to change directly on the calculations sheet if you’re so inclined.

OAS: Old Age Security payments. I think I got the default right for a couple starting at age 65. Note that the spreadsheet will not calculate clawbacks (if you make more than a certain threshold — near ~$60k/year I believe — your OAS payments will be reduced). If you will have higher income, you may want to adjust down the OAS, or build in a check for yourself in the calculations page.

Inflation will be an important factor, particularly over the length of time you’ll be retired. Officially, inflation is measured by the consumer price index (CPI), and that’s the metric many things like pensions are adjusted to. On the calculations side, it’s also the figure used to keep everything in real dollars.

However, the basket of goods used to calculate CPI may be different than the inflation rate your spending actually increases by. For instance, if you’re retired and spending most of your money on food, entertainment, energy, and utilities, then you could find your experienced inflation higher than the CPI one if those items increase in price faster than the consumer goods, clothing, and alcohol you’re not buying. Or, if you expect your spending patterns may change over time at a rate different than that of CPI: maybe as you age you’ll spend less on parties and traveling, or perhaps your medical care costs will march ever upwards. Enter your anticipated personal inflation rate in the appropriate box.

Return Rates: The rate of return for equities and fixed income is pretty self-explanatory. For equities in particular, you’ll hardly ever see the actual return exactly match the average you use here. So be sure to use a reasonable assumption for long-term returns, and don’t get swayed too much by recent events (either too high or too low).

Average Tax Rate: The tax calculation is extremely simple: just enter an average tax rate. A fudge factor is used on non-registered equity investments to indicate that they’re more tax-efficient than regular income (I simplistically used a 75% inclusion rate). The taxes are charged on and paid out of income — so if you’re burning through your TFSA or savings, the spreadsheet won’t be taking tax off that. It’s just an approximation to try to account for the fact that you’ll need after-tax income to pay your expenses, and the tax rate is user-entered and flat across all incomes and ages: it should be close enough (and doesn’t seem to be a deciding factor in any of the scenarios I’ve tinkered with). Plus, it’s a lot of bother to try to squeeze more false precision out of it. Just remember that your average tax rate is not the same as your marginal tax bracket.

The amounts in each pool of money should be self-explanatory. The number of years to keep in fixed income is for the RRSP: the tool assumes that asset allocation will be all equities except for X number of years worth of expenses kept in fixed income. For the other accounts, it’s assumed the TFSA is all equities (at the moment it’s so small I figured it was a safe assumption, but that will age the tool in a few more years), and there’s no asset allocation calculation for the non-registered accounts: it just lets you set how much goes into each, and runs from there.

All those inputs work together to give you one short output: the age you’ll be when the money runs out.

The answer to “can I retire now?” will depend on a number of factors included in the calculation, your own personal criteria, and whether that output number comes out as older or younger than your criteria figure (e.g.: your maximum expected lifespan).

Stay tuned for the next part, discussing how to use this spreadsheet to run multiple scenarios.

Most of what I write is aimed at younger folk: people my age, and also people who are likely to stumble upon and read blogs in the first place. So I talk a fair bit about investing and saving for retirement, but not much about what to do when that switch to living off your accumulated capital starts looming close.

Recently, someone outside my typical audience asked me that deceptively simple question: can I retire?

There are a fair number of calculations and factors that go into that decision. At the most basic level it all comes down to finding ways to meet your spending needs until you don’t have spending needs any more (i.e.: you die). How much will you spend, how much do you have? What will your rate of return be, and how long might you live?

People have come up with some quick rules-of-thumb which I think are a little chancy: I’m all for approximations, especially when it comes to long-range forecasting where precision is impossible, but setting some percentage withdrawal rate seems just a little too simple. So I decided to tackle the problem from something like a first principles approach: what would the finances look like after the first year of retirement — where would the money go out, where would it grow? And for the year after that? And so on.

I made a fairly simple spreadsheet that will just take some assumptions on returns, and run through the scenario of drawing down the savings while at the same time earning returns on the investments.

You can download the spreadsheet — Potato’s Retirement Calculator — right here.

It’s not the most complex retirement calculator: there are some software packages out there that include a tax package to estimate your tax liability to the dollar, and that run massive Monte Carlo simulations on a whole gamut of market return scenarios. But given how uncertain the future is, I think this should serve many of you well enough — and since it’s spreadsheet-based, it’s easy to see the calculations, fix any bugs that might be there, or customize it for your own use.

Just enter your assumptions about rates of return, inflation, pension income, CPP/OAS, and investments, and then the spreadsheet will tell you approximately how old you’ll be be when the money runs out.

There are some quirks: for one, I didn’t set it up to solve the inverse problem (given spending and life expectancy, calculate the investment nest egg needed). For another, I originally set it up for an age 60 or 65 comparison, and CPP/OAS, and pension really only fit into those ages or later. It’s not hard to change for your own case, but it’s not as automatic as the other variables.

Download it, give it a try, and let me know what you think. Also be sure to check out the next two posts for more details.

Part 2: details on variables and calculations.

Part 3: running through scenarios.

This is a monthly update from the Blueberry Portfolio. The events I talked about below happened approx 8 months ago.

Well, it’s been about a month since I started managing the Blueberry Investment Portfolio. In that time, the TSX has swung way down (12%), then most of the way back up, ending down just 2.5%.

The broker is reporting the portfolio as down just 0.3%, which I call a win, but given that the money was safely ensconced in a savings account before that, any loss can perhaps be viewed with derision.

This of course was not the case even just a few days ago. In the initial discussions of where to put the money, a planned core holding was going to be Berkshire Hathaway, the conglomerate controlled by Warren Buffett. Unfortunately I didn’t figure out how to buy US-listed securities in time, and some good news came out that drove the stock up over 10%. Though it may still be a good investment at 10% more, I didn’t feel like chasing it, and besides, buying in the US does add some complications.

Another missed opportunity was Armtec: once a darling stock in that magical “infrastructure” business that was supposed to get a boost from government stimulus spending, Armtec actually stumbled of late, and after suspending dividends started to trade like it was going out of business completely. I had just started to do my due diligence on it one Friday, planning to finish on the weekend. I didn’t get around to it that weekend, but figured I’d have plenty of time to finish over the next week. Unfortunately sometimes the market doesn’t give you the benefit of time, and by the middle of the week Armtec had jumped 120%. There’s no way to know if I would have felt comfortable buying it after doing my reading, but it still feels like a miss.

So we missed a few good opportunities.

We also had some rough luck with the stocks we did buy, finding Superior Plus melting away roughly 20% after buying it at what had already looked like bargain-basement prices. Even what I thought were solid names like Canexus and Futuremed were nearly as bad this month. The worst though was Daylight Energy, which dropped 32% over the course of the month.

However, our biggest loser turned into our biggest winner this week as a takeover offer for Daylight came through. Though we could have collected a few more percentage points by waiting until the closing date, I thought it might be prudent to sell into the market, and did just that last week, generating our first realized profit (and also bringing the portfolio almost back up to even).

And of course Daylight is an exemplary case to use in demonstrating that even when you get it all right in the market and end up making money, you certainly don’t get there by experiencing the market go straight up. We had to ride through some pretty gut-wrenching declines before we made it to today.

Speaking of take-overs, another large position in the portfolio is Capital Power Income Fund, which has received a take-over offer from Atlantic Power Corp. The deal is a little strange in that you have the option of either getting cash for your shares, or shares in the purchasing company, Atlantic Power. What’s really strange though is that about half the people will have to choose shares, or they’ll run out of cash. So odds are good that since the Capital Power shares were trading at a big discount to the take-over price, everyone will want cash. Thus, everyone is likely to end up with a mix of about half shares and half cash (though the default option is shares, so anyone who doesn’t bother to read their mail will get stuck with shares).

When we bought, Capital Power was trading as though you were just going to get shares, and at a discount to cash — but since we should get at least half as cash, we should do better than that. It works out to about a 2% return, which in just about 4 months is pretty decent (~6% annualized, with very little risk, and as much as 12% annualized if we get all cash). And that isn’t including the healthy dividend, either. Today, the Atlantic Power shares have traded lower, so even though Capital Power is still in the same place, it’s as though the discount has gone away (if we were proper arbitraguers, we would have shorted ATP).