≡ Menu

How diversification worked during the Global Financial Crisis

Last week we saw how the main asset classes are behaving during the coronavirus crisis. So far diversification is working just as if our predicament were a textbook exercise, rather than an ugly reality.

But what about the last time we went down the tubes? What asset allocation lessons can we learn by studying the Global Financial Crisis (GFC)? Did the same patterns play out and does that tell us anything about our fate this time?

Below I’ve tracked that market crash using the Exchange Traded Funds (ETFs) that were available at the time.1

All the ETF returns data and charts come from the excellent portfolio-building service JustETF. Returns include dividends but not inflation.

World equities peaked on 12 October 2007.2 They hit the bottom 17 months later on 6 March 2009.

It wasn’t until 9 March 2010 that equities climbed out of the hole – nearly two and a half years after their fall began.

100% equities during the Global Financial Crisis

World equities during the Global Financial Crisis

(The little D flags at the bottom of the graph refer to dividend payments)

The GFC was one of the worst crashes on record, but it was a fairground ride in comparison to the first month of the coronavirus slide.

In 2020 we dropped like an anvil tied to an anvil: world equities plunged 26% from 20 February to 23 March. It took a whole year for world equities to tumble that far in 2007 to 2008. After the first month, equities had only slipped 7.5%.

In 2020 there is already talk about whether we’ve missed the great buying opportunity or whether there will be more downward legs to come.

There was certainly plenty of pain ahead of us at this point in the GFC. World equities closed just 1.5% down on 19 May 2008, only to trough out at -38% on 6 March 2009. Few were predicting we’d touched the bottom, even by then.

Volatility was wild, too – drops of more than 10% in a week, as confidence drained out of the economy in October 2008.

Equity correlations ‘go to 1’ in the Global Financial Crisis

Equities go to 1 in the Global Financial Crisis

What about a diversified portfolio? How did other allocations do when everbody was making their way not-so-calmly to the exits in the GFC?

In a global crisis nowhere is safe. And so once again there was nowhere to hide in equities.

Just like this time, the financial services exposure of the London Stock Exchange hit UK equities especially hard. Commercial property took a beasting in 2007-10, too.

Emerging markets did relatively well initially, again echoing the early stages of the coronavirus crisis. But as the recession took hold from September 2008, the fearsome volatility of emerging markets dunked them down by nearly 50%.

Mind you that same volatility eventually catapulted them back. Emerging markets were more than 12% higher by 9 March 2010. The developed world was still hovering around 0%.

Emerging Markets (Orange line; ticker: IEEM)

  • Low point: -49.43%, 24 October 2008 vs World -30.2%
  • 6 March 2009: -42.32% vs World low -37.96%
  • 9 March 2010: 12.38%

Global Property (Red line; ticker: IWDP)

  • Low point: -56%, 6 March 2009
  • 9 March 2010: -8.96%

UK FTSE 100 (Blue line; ticker: ISF)

  • Low point: -44.78%, 5 March 2009
  • 9 March 2010: -9.05%

Flight to quality: government bonds in the GFC

Gilts negatively correlated with World equities during Global Financial Crisis

When we look at bonds versus equities on a bear market chart, the picture we hope to see is a mountain range reflected in a lake. When the plunging valleys of equities are mirrored by bond peaks, we know our losses are to some degree off-set.

Intermediate gilts (a blend of short, medium, and long-term UK government bonds) lived up to their safe haven billing during the GFC. You can see that gilts were broadly negatively correlated with equities for the duration of the panic. During most of the big equity sell-offs, gilts pitched in the opposite direction as investors sought refuge in safer assets. That’s the famous flight to quality.

Unlike in the early stages of the coronavirus crash, gilts didn’t turn negative, either, bar an irrelevant -0.1% day on 15 October 2007.

Unfortunately I don’t have any data stretching back to 2007 for long or short-term gilt ETFs.

However their returns would have reliably sandwiched the intermediate gilt ETF’s – long-term gilts doing even better, while short-term gilts provided only crumbs of comfort.

Intermediate UK gilts (Blue line; ticker: IGLT)

  • High point: 20.28%, 8 October 2009
  • Low point: 0.85%, 13 June 2008
  • 6 March 2009: 18.56% vs World low -37.96%
  • 9 March 2010: 16.3%

Inflation-linked gilts and corporate bonds – not so effective in a crisis

Linkers and corporate bonds during the Global Financial Crisis

Inflation-linked gilts and investment-grade corporate bonds have performed poorly relative to conventional gilts so far in the coronavirus crash, just as we’d expect.

Corporate bonds didn’t do us any favours in the dark days of 2008-09, either. That’s hardly surprising – they were backed by companies that many feared wouldn’t survive as the Great Recession laid waste to the economy.

Inflation-linked gilts (‘linkers’ to their friends) outstripped conventional gilts for the first year of the GFC. But they sold-off once deflation became the chief concern. Central banks tried to jolt the economy back into life with their QE defibrillator. That led to talk of runaway inflation, but linkers still lagged as the balance of risks suggested the patient was more likely to lapse into a coma.

Index-linked gilts (Orange line; ticker: INXG)

  • High point 1: 15.26%, 1 September 2008 vs gilts 6.72%
  • High point 2: 18.49%, 30 November 2009
  • Low point: -1.33%, 8 December 2008
  • 6 March 2009: 7.17% vs gilts 18.56%
  • 9 March 2010: 14.85%

Linkers bore up (though not as well as conventional gilts) even as the market rock-bottomed on 6 March 2009. Their inflation-fighting capabilities make them an important part of a passive investor’s load-out even though we haven’t witnessed rampant inflation in the developed world since the early ’80s.

Sterling corporate bonds (Blue line; ticker: SLXX)

  • High point: 5.67%, 18 January 2010
  • Low point: –20.46%, 26 March 2009
  • 6 March 2009: -13.93% vs World low -37.96%
  • 9 March 2010: 5.17%

Gold had a good crisis

Gold and gilts diversification during the global financial crisis

It was gold’s time to shine during the GFC – so much so I’d have been tempted to put a few of my highest carat teeth onto the market.

The yellow metal is proving to be a good diversifier during the coronavirus crisis, too.

Gold is the one accessible, low-cost asset that demonstrably traces a different path from equities and bonds. Long-term returns may be low, but it was worth its weight in 2008-09.

Gold (Yellow line; ticker: PHAU)

  • High point: 102.39%, 3 March 2010
  • Low point: 0.14%, 22 October 2007
  • 6 March 2009: 77.19%

Note that gold can be as horribly volatile as equities. It lost 20% in the week of 10-17 October 2008 while equities bounced along the bottom. Gilts didn’t help much that week either, but at least they only shed 0.5%.

Diversified portfolio vs 100% equities

A diversified portfolio during the global financial crisis

You know the drill. The 100% equities portfolio (green line) is pitched above against a diversified portfolio comprising world equities, intermediate gilts, and gold (blue line).

This diversified portfolio is:

  • 60% world equities (Ticker: IWRD)
  • 35% intermediate gilts (Ticker: IGLT)
  • 5% gold (Ticker: PFAU)

As the chart shows, when an emergency hits you’ll probably be glad you packed your parachute, inflatable arm-bands and fashion-forward face-mask. Your portfolio should be ready for anything.

Take it steady,

The Accumulator

  1. The ETFs I used for last week’s post didn’t exist in 2007. []
  2. As tracked by the iShares MSCI World ETF; Ticker: IWRD. []

Receive my articles for free in your inbox. Type your email and press submit:

{ 33 comments… add one }
  • 1 Keith April 21, 2020, 10:45 am

    One of the oddities of gold in the present crisis is that if you hold it in coins, bars, scrap or jewellery, and need to turn it into cash at the current peaks (accentuated by sterling’s weakness), it’s much harder than in normal times because so many of the usual high street dealers are closed. A couple of the Hatton Garden places appear still to be offering a restricted in-person service, but it’s much harder if you live outside London.

  • 2 Algernond April 21, 2020, 10:59 am

    For the diversified portfolio in the last chart, what is the rebalancing frequency over the period?
    Or are the %’ages shown just the weighting at the start?

  • 3 ZXSpectrum48k April 21, 2020, 11:13 am

    On linkers vs. conventional gilts. The two funds you are comparing, INXG and IGLT, had very different durations (the linker fund INXG being higher than IGLT). Looking at the performance in duration-weighted terms (or using a long duration Gilt fund rather than an intermediate duration one), the outperformance of linkers in 2008 narrows dramatically, and the underperformance over the whole GFC is much larger.

    That’s because a linker’s duration exposure is only to changes in forward expectations of real yields, while conventional gilts provide exposure to changes in forward expectations of both real yields and inflation. Obviously by the end of the GFC, expectations for both had fallen substantially, so linkers underperformed.

    Hence if you invest £1 in a linker fund, with duration of say 10, and a £1 in a conventional gilt fund, also of duration 10, then you will find the price volatility of the linker fund will be less (typically by around 30-60%, but not always). You need to have a longer duration linker fund than a normal Gilt fund to get the same bang for each £1 invested.

  • 4 gadgetmind April 21, 2020, 11:59 am

    We’re both retired, and 100% dependant on investments for income, so I’m really glad for the lessons from this site (and Mr Hale and others) regards portfolio structure. Going into this with 50% bonds in our pensions was a nice place to be.

  • 5 The Investor April 21, 2020, 1:13 pm

    @gadgetmind — You’re welcome! Glad you’re in a good place.

  • 6 Naeclue April 21, 2020, 2:50 pm

    Gilts responded well after the dot com meltdown as well, but this behaviour is not a rock solid rule. In the early 70s chaos for example, both gilts and equities lost money. Gold did well back then as well, but I will not touch gold, nor corporate bonds, which I guess makes me an active investor?

    I am sure I have said this on here before, but I do not recall a time where gilts have ever looked attractively priced. Even cash deposits have usually looked better. But it is at precisely these periods of stock market chaos when their worth is shown, usually.

    Despite this, even I (a long term holder) am starting to question the logic of holding gilts now. My shortest dated gilt matures in December 2027 and has a gross redemption yield of just 0.17%. Ordinarily I would be looking to roll that into a longer dated gilt next January. At present, a likely candidate would be the Treasury paying 4.5%, maturing in 2034, but that only has the GRY of 0.27%.

    Instead, I could lock in returns on 1,2 and 3 year FSCS protected deposits of between 1.5-1.8%. This would free up space inside tax shelters which I could move more equities into (CGT Allowance permitting, which I have not checked). I would miss out on the possibility of even better returns should gilt yields plummet even more, but to offset that I would not risk losing money should they start rising again. All getting very active, which I am generally opposed to – please talk me out of this.

  • 7 Jamie Brown April 21, 2020, 3:34 pm

    I’m very new to all of this and I am 30, looking to retire in my late 50s and have gone 100% equities with Vanguard’s FTSE Global All Cap fund, which seems fairly popular. Reading this I should probably add some bonds, at my age does that still make sense or would it be better off to hold off for a few years?

  • 8 Matt April 21, 2020, 3:46 pm

    If you have a long term time horizon you should be OK to be all equities but when you get closer to retirement move into bonds. But as this article shows bonds give you a cushion in a crisis, but outside of a crises they will limit your growth.

  • 9 Brod April 21, 2020, 4:14 pm

    @JB – Matt’s approach is very valid.

    Alternatively, you could work out what you are realistically able to save, index it to and increase with inflation and the occasional pay rise (remember them?) and say “what mix of bonds and equities is likely to get me there?” over the next 25 years. This will give you your mix and likely a smoother ride through the stock market’s occasional hissy fits.

    70/30 or even 60/40 will might well do.

  • 10 David C April 21, 2020, 4:35 pm

    @Naeclue
    You’ll need to move pretty quickly to lock in those fixed-rate savings, I think. Rates are being cut all over the place.

  • 11 Peter April 21, 2020, 5:57 pm

    > In 2020 we dropped like an anvil tied to an anvil
    Galileo would like a word.

  • 12 The Investor April 21, 2020, 7:13 pm

    @Peter — Hehe. Touche.

  • 13 ZXSpectrum48k April 21, 2020, 7:21 pm

    @Naeclue. Are you sure the UKT 4.5% 7-Dec-2034 has a yield of 0.27%? I see it at 0.48%. What about barbelling the position? If you buy £100 of UKT 4.5% 7-Dec-2034 your key parameters are: Yield 0.48% Duration 8.78, PVBP 13.03, Convexity 0.91.

    Instead put £82.40 in an FSCS deposit account at 1.5% and £17.60 in UKT 3.5% 22-Jul-2068 (Yield 0.46%, Duration 31.87, PVBP 74.061, Convexity 12.98). Combined position has following parameters: Yield 1.32%, Duration 5.60, PVBP 13.03, Convexity 2.28. So identical PVBP but higher convexity, higher yield. It has lower duration but that’s irrelevant since bond price change = PVBP x yield change, not duration x yield change (as some mistakenly think). Clear risk is obviously a steepening of the yield curve in 15s/50s .

  • 14 FIRE v London April 21, 2020, 11:17 pm

    There was a peculiar dynamic about the GFC. That was how we seriously worried about major banks going bust. I had three conversations that I remember with a chill, questioning whether $1m+ liquid sums in Goldman Sachs / Barclays / Lloyds were in fact safe. When you are worried about Lloyds / Goldmans going bust, there is nowhere obvious to move your money to – and in any case moving $1m+ took days, not hours, in those days.

    The phrase ‘too big to fail’ was invented in the aftermath. At the time it seemed a real possibility, for a week or two in Sep 2008, that other larger banks would follow Lehman Bros.

    I know banking crises are common in the history books, but in the UK no similar phenomenon has occurred in, I believe, my grandparents’ lifetime or mine.

    I think this dynamic helps explain the strong performance of Gold. Though obviously Gold does tend to well whenever panic pounds the streets.

    In this crisis, so far at least, there is no obvious worry about banking liquidity. Given that the BoE stress tests, which the banks all pass comfortably at the moment, see a fall in UK GDP of only 10%, perhaps we are being complacent.

    Every major corporate is drawing all its available credit lines right now. It will be interesting to see the banks’ figures in the next reporting season.

  • 15 Maximus April 21, 2020, 11:58 pm

    @Jamie
    I’d stay fully equity invested until you’re 5-10 years away from retiring, and then add in bonds every year until you’re happy with the mix. I stuck with 100% equities until 5 years to go, then added 10% bonds on my birthday (to keep the plan nice & mechanical) until I had 50% equities / 50% bonds, which is the mix I remain with…
    Good luck!

  • 16 The Accumulator April 22, 2020, 9:03 am

    @ Jamie Brown – So much depends on how well you handle risk once you a good chunk of your wealth is staked on the market. The danger of panicking is real. Check out some pieces on risk tolerance for more:

    https://monevator.com/how-to-estimate-your-risk-tolerance/

    https://monevator.com/what-you-need-to-know-about-risk-tolerance/

    @ ZX – thank you for those insights.

    @ Algernond – No rebalancing.

    @ Fire vs London – Yes, my memory of 2008 is that the fear factor was much greater. The End Of The World As We Know It seemed much nearer at hand. In the aftermath, you can draw a direct line to the rise of populism, austerity, Trump and, dare I say it, Brexit, so lord knows what we’ll be in for after this. Humanity United versus inequality and climate change?

  • 17 ChesterDog April 22, 2020, 9:47 am

    @Peter and TI,

    An anvil tied to an anvil falls at (approximately) the same velocity, but with a lot more momentum: it’s much harder to stop!

  • 18 Al Cam April 22, 2020, 10:30 am

    @Peter, TI, & ChesterDog
    Apparently:
    a) Galileo never actually performed the experiment;
    b) a similar experiment was actually performed before Galileo in the Netherlands; &
    c) a variation of the experiment (using a feather & a hammer) was performed on the moon in 1971!!
    see https://en.wikipedia.org/wiki/Galileo%27s_Leaning_Tower_of_Pisa_experiment

  • 19 Al Cam April 22, 2020, 10:39 am

    @ZX, re barbelling:
    Assuming there is no short-term need for the funds in question, I thought such an approach worked best in a practical sense in a rising interest rates environment, as you periodically re-new the short term money to take advantage of the rising rates. However, I am not sure I can see rates increasing any time soon. And, I guess rates may go the other way.
    Or, have I got this wrong?

  • 20 Grislybear April 22, 2020, 10:52 am

    @ZXSpectrum48k. Your suggestion for a barbell using cash and long bonds is excellent. It’s a combination that’s very appealing ,thanks for sharing.

  • 21 Naeclue April 22, 2020, 12:40 pm

    @ZXSpectrum48k, quite right. I had the coupon as 4.25% – hazards of copy and paste. Based on a new price for UKT 4.5% 7-Dec-2034, I obtained today of 156.3 for settlement 26/04, I am getting a closer yield to your 0.48% of 0.45%, but your sensitivity figures don’t look right (Duration 8.78, PVBP 13.03, Convexity 0.91.). I am getting Duration 11.52, PVBP 18.02, Convexity 0.79. For a bond maturing in roughly 14.5 years, the coupon rate would have to be much higher than 4.5% in order to bring the duration down to 8.78. More like 20%.

    I have been considering a bar-bell approach as well though, but through a bond ETF and more in cash, rather than messing about with individual gilts. IGLT and VGOV now have TERs of only 0.07%, effective duration about 13, and 14.6 respectively. It would be much easier for my wife to deal with should she outlive me, or I become incapable (or screw up my cut and pasting ;))

  • 22 Neverland April 22, 2020, 12:49 pm

    @FireVLondon

    ‘I know banking crises are common in the history books, but in the UK no similar phenomenon has occurred in, I believe, my grandparents’ lifetime or mine’

    Last banking crisis prior to 2008 was in the mid-70s, so every 25-30 years

    Always the same: deregulate; and then repent

    But back then the code of the city was omerta

    https://www.ft.com/content/e5679bd6-c096-11dd-9559-000077b07658

    The 70s banking crisis came off the back of a stock market crash, but this doesn’t feel like the same thing, because banks (but not shadow banks) are still strongly regulated after the last crisis

  • 23 ZXSpectrum48k April 22, 2020, 12:52 pm

    @Naeclue. I think messing around with individual bonds is possibly more trouble than it’s worth. In fund terms, rather than buy say an intermediate duration Gilt ETF like IGLT (duration 12.9) you could instead buy £60 of a long duration Gilt tracker like VUKLDGI (duration 21.5). You maintain exactly the same duration exposure but use £40 less cash, which can go into the deposit account. Obviously this has gives you a curve steepening exposure between 0 and 12.9 years and a flattening exposure, between 13 and 21.5 years in duration terms, so it’s no free lunch but it does allow you to get some yield if that is your aim.

    Regarding the Gilt, rereading my comment I think you’re correct. I’ve got a feeling at some point I switched in my sheet from the 2034 bond to the benchmark 10-year Gilt 4.75% Dec-2030 since the duration/convexity is wrong. I really shouldn’t watch Star Wars IX while watching fiddling with sheets. The principle is the same though. Apologies.

  • 24 Naeclue April 22, 2020, 1:00 pm

    I remember well that scary time in 2008. We had a considerable sum in a NatWest business account that I had to very rapidly reduce. A combination of large pension contributions (annual allowance in those days was over 200k), in advance salary payments, purchase of a short dated gilts ETF and even loans to people I could trust who temporarily paid the money into NS&I. Took out more accounts across banks and got the loans back by company year end, so accountant was happy.

    I hope we never see anything like that again but I still avoid having more than the FSCS limit in any bank. NS&I are good with their £1m limit per person.

  • 25 Naeclue April 22, 2020, 1:15 pm

    @ZXSpectrum48k, My SIPP provider (HL) charges an extortionate platform fee for OEICs, so I tend to steer clear of them. I could purchase in my ISA with iWeb though as they do not have that problem.

  • 26 Naeclue April 22, 2020, 1:31 pm

    Short term rates just seem to keep on dropping. I have a 3 year NS&I growth bond maturing in October, currently paying 2.2%, I think I will be lucky to get 1.5% by then on a rollover (should they even offer it). One other development I don’t like is that these term deposits have universally become what they say. Invest for 3 years and there is no early access at all (maybe on death). Previously you could get access with say a 90 day interest penalty should you need to, but even NS&I have stopped that.

  • 27 Al Cam April 22, 2020, 1:38 pm

    Re 2008 debacle:
    almost nobody knew there was such a thing as the FSCS (or whatever it was known as then) protection limit. I even have text books that predate 2008 that describe cash variously as safe, risk free, etc.

  • 28 Neverland April 22, 2020, 1:47 pm

    @Naeclue

    NS&I is the unwanted step child from the first marriage of government backed savings products

    LISA is the golden child

    Its a shame, NS&I index linked certificates were a really good product – still have some

  • 29 Al Cam April 22, 2020, 1:49 pm

    @Naeclue:
    re “short term rates ….”
    Agree 100% with all you say.
    What I have taken to doing is viewing any interest paid as a bonus versus my nominal assumed annual inflation target – which I have set based on >100 years of RPI history.

    BTW, I do vaguely recall something about some deposit takers permitting early access to fixed rate accounts at the outset of the current C-19 episode – but cannot quickly find any relevant links.

  • 30 Neverland April 23, 2020, 9:20 am

    Well lots of gilts to buy now apparently £180bn to be sold over the next few months to July

    National debt now over 100% of GDP

    No doubt all is fine

    https://www.dmo.gov.uk/media/16478/sa230420.pdf

  • 31 David C April 23, 2020, 3:13 pm

    @Naeclue It may not help in your situation, but fixed-rate Cash _ISAs_ do normally allow early withdrawals (some only allow complete closure), albeit with a x-days interest penalty. I assume that this must be a rule. At the mo though, they generally pay a little less than their non-ISA equivalents.

  • 32 Naeclue April 23, 2020, 4:58 pm

    @Neverland, yes and a fair sized chunk will be taken up by the BoE, swapped for pound notes. I guess the question then is where will those pound notes go? Let’s hope that a large proportion of holders don’t decide they would prefer dollars instead.

  • 33 MrOptimistic April 24, 2020, 12:42 am

    My problem is that I don’t know how to deal with bonds. I am inclined to think they are not the low risk asset inversely correlated to equities they used to be.

Leave a Comment