Latest / How Canadian Markets Work / Episode 41: Duration

Transcript

Episode 41: Duration · How Canadian Markets Work

0:00
OK, so I want you to imagine a scenario that,
0:02
honestly, it sounds like a glitch in the matrix.
0:04
Oh, a glitch. Yeah, total glitch. Right. So picture
0:07
two bonds. Both of them are government quality,
0:10
meaning there's practically zero risk that the
0:14
issuer is actually going to default. Right, the
0:16
safest of the safe. Exactly. And both of them
0:18
pay a really respectable 5 % coupon. To the average
0:23
investor, I mean, they look like identical twins.
0:25
They look like the safest possible place you
0:27
could park your cash. But then, interest rates
0:30
suddenly rise by exactly 1 % across the board.
0:33
Which is, I mean, that's a completely normal
0:34
economic event. Central banks adjust rates all
0:37
the time to, you know, cool inflation or stimulate
0:40
growth. Totally normal. But the reaction in this
0:43
specific scenario, not normal at all. The first
0:47
bond falls in value by a little under 3%. It's
0:50
a minor dip. You barely even feel it. But the
0:53
second bond... the identical twin, it plummets
0:56
by more than 11 .5%. Oh wow, that's substantial.
0:59
Right, same market, same interest rate hike,
1:02
same 5 % coupon, same rock -solid government
1:05
backing. So how can two safe bonds behave so
1:10
drastically, wildly differently. That is the
1:12
million dollar question. It really is. And to
1:15
figure out why this happens, we're doing a deep
1:18
dive today into a really fascinating stack of
1:21
excerpts on Canadian market mechanics. Specifically,
1:24
we're focusing on the structural drivers of bond
1:27
price volatility. And it really is the ultimate
1:29
blind spot for investors. You know, you go into
1:31
the fixed -in -queue market looking for safety,
1:33
and you just accidentally buy this massive exposure
1:36
to interest rate changes. Yeah. So the mission
1:39
for this deep dive is to decode the single most
1:41
important and frankly the most widely ignored
1:44
number on every single bond fund fact sheet in
1:47
existence. The one nobody looks at. Exactly.
1:49
I mean, it takes exactly four seconds to look
1:50
it up, but almost nobody outside the financial
1:53
industry ever checks it. And that number is called
1:55
duration. Yes, duration. Because people buy bonds
1:59
specifically to avoid volatility, right? They
2:01
want out of the stock market. But if you don't
2:03
know what we're talking about today, you might
2:05
literally be buckling yourself into an absolute
2:07
freefall without even realizing it. Duration
2:09
is the hidden key here. To understand why one
2:12
of our identical twin bonds fell four times as
2:15
hard as the other one, we have to look at what
2:18
duration actually measures. And I will warn you
2:21
up front. It has this sort of dual identity.
2:23
Oh, really? Yeah. It has two definitions that
2:25
initially sound like they're talking about completely
2:27
different things, even though they are mathematically
2:30
the exact same concept. OK, let's unpack the
2:32
first definition then, because I think this is
2:34
the one that causes the immediate pain for people.
2:37
Yeah, the pain. So definition one. Duration is
2:41
a pain multiplier. It basically measures a bond's
2:45
price sensitivity to interest rate changes. Exactly.
2:48
It is a pure sensitivity metric. So if a bond
2:51
has a duration of five, its price will move roughly
2:54
5 % for every 1 % change in yield. And you have
2:58
to remember the golden rule of bonds, right?
3:01
Prices and yields always move in opposite directions.
3:03
Right, they're on a seesaw. Exactly. So if overall
3:06
interest rates go up by 1%, your bond with a
3:10
duration of five goes down by roughly 5 % in
3:13
price. Got it. If the duration is 10, you drop
3:15
10%. It just tells you exactly how much an interest
3:19
rate move is going to hurt you. Simple enough.
3:22
I mean, I multiply the rate change by the duration.
3:25
But then you read the textbook, and you get definition
3:27
number two. Right. And definition two is duration
3:30
is the weighted average time until you get your
3:32
money back. Where each future payment is weighted
3:35
by its present value. Yes. And when you hear
3:37
that for the first time, your brain kind of just
3:40
short circuits. Because how is a percentage multiplier
3:44
of pain the exact same thing as a timeline of
3:46
when I get my cash? I know it sounds totally
3:48
disconnected, but what's really fascinating here
3:51
is how deeply connected those two ideas actually
3:54
are. Okay, lay it on me. The longer you have
3:56
to wait to get your money, the more sensitive
3:58
its present value is to whatever current interest
4:01
rates are doing. We have to think about compounding
4:04
here. Okay, compounding. Yeah. So if I owe you
4:07
$100 tomorrow, and the central bank raises interest
4:10
rates today, you don't really care. No, because
4:13
it's just tomorrow. I'm getting paid in like
4:15
24 hours. Right. But if I owe you $100 30 years
4:18
from now and interest rates rise today, the present
4:21
value of that distant payment absolutely collapses.
4:25
OK, wait, let me make sure I'm following this.
4:27
The collapse happens because if rates are higher
4:29
today. I could take a much smaller amount of
4:32
money right now, put it in a new bank account
4:34
with that new higher interest rate, and just
4:36
let it grow for 30 years to reach $100. That
4:39
is the core of it, yes. Because of the power
4:42
of compounding... A small change in interest
4:45
rates today amplifies exponentially over three
4:48
decades. Oh, I see. The new bonds hitting the
4:52
market, they are paying that higher rate for
4:54
30 whole years. So your old bond, the one that's
4:58
stuck at the lower rate, it becomes massively
5:00
less attractive. Right. Nobody wants the old.
5:03
worse rate. Exactly. So the only way to convince
5:06
someone to actually buy your old bond is to slash
5:09
its price. And the further out that final payment
5:12
is, the deeper the price cut has to be. OK, let
5:16
me try a physical analogy to just lock this in.
5:18
Let's hear it. Think of holding a really heavy
5:21
weight like a 10 pound dumbbell on the end of
5:24
a stick. OK. And the stick represents time. If
5:28
you hold that weight on a really short stick,
5:30
say like one foot long, you can hold it steady.
5:33
your wrist feels it, but it's totally manageable.
5:36
Right. But if you take that exact same 10 -pound
5:38
weight to the end of a 10 -foot pole and you
5:40
try to hold it straight out, I mean, the leverage
5:42
is just enormous. The weight itself hasn't changed
5:45
at all, but it feels infinitely heavier on your
5:47
wrist. I love that. Building on that analogy,
5:50
the longer the stick, so the longer you wait
5:52
for your cash, the more leverage the interest
5:54
rate has to swing the price of your bond. OK.
5:56
The timeline inherently is the sensitivity. The
5:59
stick itself is the multiplier. That makes so
6:01
much sense. So if time is the ultimate lever
6:04
dictating this sensitivity, we really need to
6:07
look under the hood here. We do. What structural
6:10
features of a bond actually manipulate that timeline?
6:15
Like what makes the stick longer or shorter?
6:18
It seems to me like the maturity date would be
6:19
the obvious starting point. Oh, maturity is absolutely
6:22
the dominant factor. The maturity date is simply
6:24
when the bond ends and the principal is finally
6:26
repaid. Right. So a longer maturity date automatically
6:30
pushes out the timeline, which dramatically increases
6:32
duration. A 30 year bond inherently has a much
6:36
longer duration than a two year bond because
6:38
the bulk of your return of getting your original
6:41
investment back is pushed decades into the future.
6:44
That makes intuitive sense. Pushing the biggest
6:45
payment decades into the future makes the stick
6:47
much longer. Exactly. But maturity can't be the
6:50
only thing dictating time, right? Yeah. What
6:53
if a 30 -year bond pays me a really massive chunk
6:56
of cash every single year along the way? Doesn't
6:59
that change the timeline? It changes it significantly,
7:01
and that actually brings us to the second structural
7:04
driver, which is the coupon size. Okay. A higher
7:07
coupon actually shortens your duration. Wait,
7:10
wait, why does getting a higher coupon shorten
7:11
the duration? I would think more money means
7:14
more risk or something. Well, think about when
7:17
that money actually arrives. A high coupon means
7:19
a bigger chunk of your total return is being
7:21
handed back to you sooner, right in those early
7:24
years. You are getting cash in hand today, next
7:28
year, the year after. So that shrinks the weighted
7:31
average time you are left waiting for your total
7:34
cash. Ah, I see. OK, so going back to my physical
7:37
analogy with the stick. Pulling those payments
7:40
forward is like sliding that 10 -pound weight
7:42
further down the stick closer to your wrist.
7:45
Exactly. The pull itself might still be 10 feet
7:47
long, but the center of gravity has shifted closer
7:49
to me, which gives the interest rate less leverage
7:52
to swing the price around. That is spot on. A
7:55
high coupon bond pays you back a significant
7:57
portion of your investment before the final maturity
8:00
date. And because you're getting your money back
8:02
faster, your overall sensitivity to interest
8:05
rates drops. Right. The cash flow basically acts
8:08
as a counterweight to the maturity date. OK,
8:10
so we have maturity pushing the timeline out,
8:13
making it more sensitive, and large coupons pulling
8:16
the timeline in, making it less sensitive. Yes.
8:20
Third factor here. The third driver is the overall
8:23
yield environment itself. The environment. Yeah.
8:25
So when general market yields are high, the present
8:28
value of distant payments is discounted much
8:30
more heavily. That means the earlier payments
8:33
you receive carry more weight in the calculation,
8:35
which slightly shortens duration. Oh, OK. But
8:38
honestly, for practical purposes, if you just
8:40
want to understand what's really moving the needle,
8:42
maturity and coupon size are the two main forces
8:45
you need to watch. So if getting cash back early
8:48
via coupons lowers your duration, what happens
8:51
if we take that to the absolute limit? Like,
8:53
what if we strip the coupons away entirely? Oh,
8:55
well, then you have reached the absolute peak
8:58
of sensitivity. You have built what's called
9:00
a zero -coupon bond. Zero -coupon bond. And it
9:03
is exactly what it sounds like. It pays absolutely
9:05
nothing. No semiannual interest, no cash flow
9:08
along the way, nothing until the very final maturity
9:11
date when you get the principal back in one big
9:14
lump sum. Wow. OK, so because there are no early
9:17
payments, pulling that average time down, the
9:20
stick is as long as it can possibly be. Like
9:23
all of the weight is perfectly balanced on the
9:25
very tip of the pole. Exactly. For a zero coupon
9:28
bond, its duration perfectly equals its maturity.
9:32
So a 20 -year zero coupon bond has a duration
9:35
of exactly 20. There is literally no counterweight.
9:39
But why would anyone buy a financial instrument
9:41
that is maximizing its own vulnerability like
9:44
that? I mean, it sounds incredibly dangerous.
9:46
Is this just some theoretical academic concept
9:48
or do people actually use these things? Oh, they
9:51
are heavily utilized, but mostly by professionals.
9:53
OK. This specific structure makes zero coupon
9:56
bonds the most interest rate sensitive conventional
9:59
instruments in existence. Wow. And savvy institutions
10:02
like pension funds or big macro hedge funds,
10:05
they use them deliberately when they want maximum
10:07
exposure to rate changes. So they want the risk.
10:10
Exactly. Think about it. If you are a fund manager
10:13
and you have a really strong conviction that
10:15
central banks are about to slash interest rates,
10:17
you want the longest duration possible. Because
10:19
when rates fall, prices rise. Yes. A zero coupon
10:24
bond gives you maximum upside leverage without
10:27
even having to use margin or any complex derivatives.
10:31
OK, I get the upside. But if you're wrong and
10:34
rates actually go up, it gives you maximum pain.
10:37
The leverage. cuts both ways, brutally, which
10:40
is why the Canadian market excerpts we're looking
10:42
at focus so heavily on these structural drivers
10:44
of volatility. When the Bank of Canada you know,
10:47
adjust rates, the ripple effect isn't just uniform
10:50
across the whole bond market. The zero coupon
10:53
bonds and the long duration bonds, they experience
10:55
massive price swings. While the short duration
10:58
bonds, they barely move at all. OK, let's put
11:01
some actual concrete numbers to our original
11:03
mystery to prove this isn't just theory. Let's
11:05
look at this volatility gap in action. Let's
11:07
do the math. So we had two identical twin bonds.
11:10
Both are government backed. Both are priced at
11:12
one thousand dollars initially. And both are
11:14
yielding 5%. Right. And suddenly, interest rates
11:17
rise by exactly 1%. So they move from 5 % to
11:20
6%. OK, so bond number one is a three -year bond.
11:23
Because it matures quickly, its duration is roughly
11:26
2 .7. So when rates rise to 6%, the price of
11:30
that bond falls from $1 ,000 to roughly $973.
11:34
So it drops about 2 .7%. Exactly. Which matches
11:38
the duration number almost perfectly. A 1 % rate
11:41
hike times a 2 .7 duration equals a 2 .7 % price
11:45
drop. The math is very clean there. Now look
11:48
at bond number two. It's a 20 -year bond. Because
11:51
it takes two decades to get the principal back,
11:53
its duration is much longer, roughly 12 .5. Right.
11:57
So when rates hit 6%, that bond falls from $1
12:00
,000 to roughly $885. So it drops about 11 .5%.
12:05
And again, a 1 % hike multiplied by a 12 .5 duration
12:09
gets us right in that exact ballpark. The math
12:11
holds up. It does. But now, let's look at the
12:14
extreme case we discussed earlier, the 20 -year
12:17
zero coupon bond. Oh, this one is rough. So it
12:20
pays a 5 % yield, but because it pays literally
12:23
no coupons along the way, you buy it at a massive
12:26
discount today. Instead of $1 ,000, it's initially
12:30
priced at roughly $376. And its duration is exactly
12:33
20, because it's a 20 -year zero. Yep. So when
12:37
the required yield in the market rises to 6%,
12:40
the price of that exact same zero coupon bond
12:44
falls to about $311. I am looking at these percentages
12:48
right now. And that is a drop of over 17 % on
12:53
a single 1 % interest rate move. It's massive.
12:57
Someone could buy a government -backed 20 -year
13:00
zero coupon bond thinking they are being incredibly
13:04
conservative because they read government bonds
13:06
are safe. Right. And they could instantly lose
13:08
17 % of their portfolio without the issuer defaulting,
13:12
without the government going bankrupt, literally
13:14
without anyone doing a single thing wrong. The
13:16
bond functioned exactly as designed. The math
13:19
is just ruthless. If you don't understand duration,
13:22
you really don't understand the risk you were
13:24
holding. You thought you were buying safety,
13:26
but you actually just bought a highly leveraged
13:28
bet on interest rates. Oh, wait, I'm looking
13:30
at the math here, and something is a little fuzzy
13:33
to me. Yeah. OK, what's fuzzy? You said the 20
13:35
-year bond had a duration of 12 .5, and it fell
13:39
11 .5%. And then the zero coupon bond had a duration
13:42
of 20, and it fell about 17%. Right. The multiplier
13:46
is close, obviously, but it's not absolutely
13:49
perfect. I mean, a duration of 20 implies a full
13:52
20 % drop, not 17%. Is duration just like a bad
13:56
estimate? Oh, no. It is a phenomenal estimate
13:59
for f— Small moves like, you know a quarter of
14:02
a percent rate hike, but you've hit on a really
14:04
crucial nuance here The relationship between
14:07
a bond's price and its yield isn't just a straight
14:09
line in finance We call the actual shape of that
14:12
relationship Convexity. Convexity. OK, break
14:15
that down for me. Why is it just a straight line?
14:17
Think about how price drops compound over time.
14:20
As a bond's price falls lower and lower, a flat
14:23
1 % rate increase represents a smaller and smaller
14:26
relative hit to its remaining value. Oh, I see.
14:28
So the curve starts to actually flatten out.
14:30
And because of this curve, this convexity, duration
14:33
slightly overestimates the price drop when yields
14:36
rise, and it actually underestimates the price
14:38
gain when yields fall. Wait. So the amount actually
14:41
curves in the investor's favor. Yes. Because
14:44
of convexity, a bonds price actually rises more
14:47
when yields fall than it drops when yields rise
14:51
by the exact same amount. It sounds like a built
14:53
-in shock absorber. It acts very much like a
14:55
shock absorber, yeah. And it's exactly why institutional
14:58
fixed income managers completely obsess over
15:00
convexity in their portfolios. Right. But for
15:03
everyday purposes, you really just need to know
15:05
that duration is your baseline measurement. Convexity
15:09
is just the secondary force that bends the math
15:11
slightly in your favor during those really large
15:14
rate swings. Okay. That helps clarify things
15:17
a lot. Now, speaking of labels and terms, I also
15:21
see different terms for duration when I look
15:23
at fun sheets, like I'll see Macaulay duration,
15:26
modified duration, effective duration. Do we
15:29
need to just pull our hair out over the differences
15:30
between these? I strongly advise against agonizing
15:33
over the labels, honestly. But it does help to
15:36
know what they represent. OK, give me the quick
15:38
version. So Macaulay duration measures the timeline
15:40
in actual years. It's the literal physical length
15:43
of the stick we talked about. OK. Modified duration
15:46
translates that timeline into the pure percentage
15:48
sensitivity metric we've been using to calculate
15:51
the price drops. Ah. So Macaulay is the time,
15:54
and modified is the pain multiplier. Exactly.
15:57
And effective duration duration just adjusts
15:59
the math for bonds that have tricky embedded
16:02
options, like if a corporation is allowed to
16:05
pay you back early. Got it. But for 99 % of situations,
16:10
whatever duration number the mutual fund or the
16:13
ETF publishes on its fact sheet is just the number
16:15
you multiply by the interest rate change. Don't
16:18
overcomplicate it. I do have to ask a sort of
16:20
boundary question here though. Just to be totally
16:23
clear, duration acts like a shield or maybe a
16:26
warning label for interest rates. But it says
16:29
absolutely nothing about credit risk, right?
16:31
Like if I buy a corporate bond with a super short
16:33
duration of two and the company just goes bankrupt
16:36
tomorrow, my short duration doesn't save me at
16:38
all, does it? No, and that is a fatal mistake
16:40
people make. Duration isolates one specific risk,
16:44
and that is interest rate risk. It assumes the
16:46
company or the government is actually going to
16:48
pay you. If the company defaults, your duration
16:52
number is completely irrelevant because the cash
16:54
flows themselves have just vanished. Wow. Yeah.
16:56
You cannot evaluate bankruptcy risk with an interest
16:59
rate metric. They are totally separate. So if
17:02
duration isolates this massive hidden interest
17:05
rate risk, how does the everyday investor actually
17:08
use this single number? to protect themselves.
17:11
What is the practical application of this entire
17:13
deep dive? It all comes down to the golden rule
17:16
of fixed income investing. match a bond fund's
17:20
duration to your actual time horizon. Your time
17:23
horizon is your true risk capacity. Let's put
17:26
a real world scenario to that golden rule just
17:28
to ground it. Let's say you are saving up for
17:31
a house deposit. Okay, perfect example. And you're
17:33
gonna buy a house in exactly three years. You
17:35
don't want your money in the stock market because
17:37
it's way too volatile. So you look up a bond
17:40
fund, it's heavily marketed as a... conservative
17:43
or a core bond fund and you just dump your house
17:46
deposit in there. But you look at the fact sheet
17:48
and that specific fund actually has a duration
17:51
of 12. If you do that, you are taking on intense
17:55
interest rate risk that you have absolutely no
17:58
capacity to wait out. Because you need the money
18:00
in three years, not 12. Exactly. If rates rise
18:03
by just 2 % next year, that fund drops by roughly
18:07
24%. Ouch. You need the money in three years.
18:10
The fund simply won't have time to recover through
18:13
the new higher yield payouts. You will literally
18:17
be forced to sell at a massive loss just to close
18:20
on your house. So despite the nice marketing
18:22
label, it wasn't conservative for you at all?
18:25
Not at all. Whenever you see a bond fund marketed
18:27
with words like safe, income, or conservative,
18:30
you must verify it by checking the duration.
18:33
Right. Marketing labels are just designed to
18:35
sell funds. Duration is the arithmetic reality
18:37
of what you are actually buying. If the duration
18:40
is long, it isn't truly conservative for a short
18:42
-term need. It has simply exchanged credit risk
18:45
for intense interest rate risk. That is such
18:48
a vital distinction to make. using a high -duration
18:51
fund for a short -term goal, it's like using
18:55
a sledgehammer to hang a really small picture
18:57
frame. Yes. Like, yes, it's a tool and it works
19:00
great for breaking up concrete, but it's the
19:02
completely... wrong tool for this specific job
19:06
and you're just going to destroy the wall. Bonds
19:08
aren't just universally safe, they are finely
19:11
calibrated tools. You have to know exactly how
19:14
heavy the sledgehammer is before you swing it.
19:16
And the duration number gives you the exact weight
19:18
of the hammer. So bringing it all together, duration
19:20
is your pain multiplier. It tells you exactly
19:23
how much your bond or your bond fund will fall
19:25
if interest rates rise. It's driven by how long
19:28
you have to wait to get your money back, which
19:29
is basically manipulated by maturity dates, pushing
19:33
the timeline out and large coupon payments pulling
19:37
the timeline in. Spot on. A longer wait means
19:39
a bigger drop and your ultimate defense is just
19:42
matching that duration to the exact timeline
19:45
of when you actually need the cash. It really
19:47
is the ultimate defense mechanism against rate
19:50
volatility, but this actually raises a really
19:52
important, maybe slightly provocative question
19:55
to consider as we wrap up today. Okay, let's
19:57
hear it. We established that duration only measures
20:01
interest rate risk, not credit risk. Right, the
20:04
bankruptcy problem. Exactly. Imagine an investor
20:08
who is absolutely terrified of interest rates
20:11
going up. They go out and buy a high -yield bond
20:14
fund, which is commonly known as a junk bond
20:16
fund. And they buy it because it boasts a wonderfully
20:19
short duration of, say, 2 .0. Are they accidentally
20:23
tricking themselves? Oh, wow. Yeah, they probably
20:26
feel completely invincible against interest rate
20:28
changes. Like, if rates rise 1%, they only drop
20:31
2%. They probably think they've just hacked the
20:33
whole system. But they're completely ignoring
20:35
the fact that high -yield bonds are issued by
20:38
fragile, highly -indebted companies. If the overall
20:41
economy falters, massive default risks could
20:44
wipe out their capital entirely. The short duration
20:48
won't protect them from the underlying businesses
20:50
failing. It's a stark reminder that solving one
20:53
variable in finance doesn't magically eliminate
20:55
the rest of the equation. That is a phenomenal
20:58
point to leave people with. You completely fix
21:01
the interest rate risk, but you accidentally
21:03
blow up your credit risk in the process. It goes
21:06
all the way back to our two identical twin bonds
21:09
at the very beginning of this deep dive, 5 %
21:11
coupons. On the surface, they look like the exact
21:14
same promise. But underneath, the plumbing was
21:17
entirely different. One was a simple speed bump
21:20
and the other was a trap door. Check the duration,
21:23
match your timeline, and honestly, just don't
21:25
let the marketing labels do your thinking for
21:26
you.