Stocks for the long run?

[Duke of Marmalade]

Besides, can you imagine what a spike in borrowing costs of that order would do to equities?

After a bit of Googling, I see that the negative correlation of bonds with equities is accepted wisdom and has itself led to a demand for bonds. But I do think you are massively overrating the syndrome. YTD it has yielded you a 2% kicker compared to say 0% on cash.
But consider the above line extracted from your recent post. I presume you mean that if bond yields and therefore interest rates spiked to 2% equities would take dive - the total reverse of negative correlation.
There is no telling the macroeconomic fall out of this unprecedented crisis but it is not too far fetched to imagine a big spike in bond yields as governments everywhere scramble to fund their deficits.
I stand by my assertion that at these yields bonds present a very asymmetric risk reward proposition for the retail investor, indeed aggravated by the fact that a big spike in borrowing costs would hit equity markets hard as well.
To me this is wait and see territory, far too much uncertainty, cash is king.

 

[Sarenco]

Duke,

Compare the performance of the following two portfolios over the last 40 years -

Portfolio 1 comprised 50% US equities (total market) and 50% cash (3-month T-Bills), rebalanced annually. Annualised return of 7.87%, worst year -17.75%; maximum drawdown -26.83%.

Portfolio 2 comprised 50% US equities (total market) and 50% long-term (30-year) Treasuries, rebalanced annually. Annualised return of 10.68%; worst year -7.26%; maximum drawdown - 20.39%.

So, the portfolio with long-term treasuries had a materially higher return, with lower drawdowns.

Why? Diversification.

Portfolio 1 was 100% correlated with the US stock market. As you said yourself, the correlation between equities and cash is exactly zero. Cash diluted equity risk but provided zero diversification benefit.

Portfolio 2 was only 80% correlated with the US stock market. Replacing cash with long-term treasuries actually created a more efficient portfolio from a risk/reward perspective.

Now, cash may well outperform long-term treasuries over the coming decades (although I personally think that is highly unlikely).

But cash cannot, by definition, diversify an equity portfolio.

 

[Duke of Marmalade]

Duke,

Compare the performance of the following two portfolios over the last 40 years -

Portfolio 1 comprised 50% US equities (total market) and 50% cash (3-month T-Bills), rebalanced annually. Annualised return of 7.87%, worst year -17.75%; maximum drawdown -26.83%.

Portfolio 2 comprised 50% US equities (total market) and 50% long-term (30-year) Treasuries, rebalanced annually. Annualised return of 10.68%; worst year -7.26%; maximum drawdown - 20.39%.

So, the portfolio with long-term treasuries had a materially higher return, with lower drawdowns.

Why? Diversification.

Portfolio 1 was 100% correlated with the US stock market. As you said yourself, the correlation between equities and cash is exactly zero. Cash diluted equity risk but provided zero diversification benefit.

Portfolio 2 was only 80% correlated with the US stock market. Replacing cash with long-term treasuries actually created a more efficient portfolio from a risk/reward perspective.

Now, cash may well outperform long-term treasuries over the coming decades (although I personally think that is highly unlikely).

But cash cannot, by definition, diversify an equity portfolio.

This helps pass the time
Back down the rabbit hole. I accept that cash does not bring (significant) diversification benefit but that is not because it has zero correlation, it is because it has zero risk. Diversification in a portfolio is when the standard deviation of the portfolio return is less than the weighted sum of the standard deviations of the components. As a mathematical tautology since the standard deviation of cash is effectively zero there is no diversification effect. But this is not a function of zero correlation. See the following graphic

The yellow line is the rolling 5 yr correlation of US bond returns and equity returns over the last 40 years. It can be seen that for the first (inflationary) period the correlation is positive (a bad thing) but in the low inflation of recent times the correlation has indeed been negative (a good thing). Overall it looks around zero correlation over the 40 year period. But zero correlation can still give significant diversification benefit.
Jumping out of the rabbit hole
I have already said that I fully bought into the role of bonds in a balanced portfolio in normal times. But nothing can convince me of any role for bonds with negative yields.

 

[Sarenco]

I accept that cash does not bring (significant) diversification benefit but that is not because it has zero correlation, it is because it has zero risk

With respect Duke, that really is indulging in semantics.

Adding cash to a portfolio of stocks in Portfolio 1 didn't add any diversification benefit because the overall portfolio was at all times 100% correlated to the US stock market. Cash cannot "respond" (positively or negatively) to stock market movements.

The yellow line is the rolling 5 yr correlation of US bond returns and equity returns over the last 40 years. It can be seen that for the first (inflationary) period the correlation is positive (a bad thing) but in the low inflation of recent times the correlation has indeed been negative (a good thing). Overall it looks around zero correlation over the 40 year period. But zero correlation can still give significant diversification benefit.

Absolutely. In fact, I think there may have been a modestly positive correlation between stocks and bonds over the full time period.

However, the correlation turned sharply negative at the most opportune times (2000, 2008, last month).

 

[Duke of Marmalade]

With respect Duke, that really is indulging in semantics.

Adding cash to a portfolio of stocks in Portfolio 1 didn't add any diversification benefit because the overall portfolio was at all times 100% correlated to the US stock market. Cash cannot "respond" (positively or negatively) to stock market movements.

Absolutely. In fact, I think there may have been a modestly positive correlation between stocks and bonds over the full time period.

However, the correlation turned sharply negative at the most opportune times (2000, 2008, last month).

I concede you have won the diversification argument. But still I can see no justification for investing in bonds at negative yields.

 

[Steven Barrett]

 

[Sarenco]

Yup, Steven's chart shows that the Agg outperformed the S&P500 over the last 20 years.

I think it's interesting to compare the performance over the last 40 years of an all-equity portfolio (S&P500) with a traditional portfolio made up of 60% equities (S&P500) and 40% bonds (10-Year Treasuries) -

1. 100% equities - annualised return 10.92%; worst year -37.45%; maximum drawdown -50.97%
2. 60/40 portfolio - annualised return 10.25%; worst year -14.00%; maximum drawdown -26.46%.

So the all-equity portfolio modestly outperformed the 60/40 portfolio over the full 40-year period but that outperformance came with some stomach churning drawdowns.

 

[Steven Barrett]

Ran figures for Global Bonds and MSCI All Country which European investors may use a bit more. Equities still lag behind bonds over the last 20 years. Investment period is a lot shorter than the S&P500 stats put up previously

 

[lledlledlled]

Ran figures for Global Bonds and MSCI All Country which European investors may use a bit more. Equities still lag behind bonds over the last 20 years. Investment period is a lot shorter than the S&P500 stats put up previously

View attachment 4416

Am i missing something here? I thought equities have out-performed all other asset classes over almost any reasonably long-term time frame?

 

[cremeegg]

Am i missing something here? I thought equities have out-performed all other asset classes over almost any reasonably long-term time frame?

Seems that is not true.

While Property returns are difficult to substantiate, I suggest that make all the above look, rather poor.

 

[Sarenco]

I thought equities have out-performed all other asset classes over almost any reasonably long-term time frame?

You thought wrong. That's the key takeaway from this thread.

 

[Duke of Marmalade]

Steven those are interesting stats. But they are history.
Equities earned 3.03% p.a. over the last 20 years. They may well do that again over the next 20 years. They might do 6% p.a. or they might go sideways.
Bonds earned 4.25% p.a. over the last 20 years. With yields on long term bonds currently at 0%, near as makes no difference, it would be in defiance of the law of gravity for them to earn 4.25% over the next 20 years. I say to anybody uninitiated in these matters that those historic returns might give a flavour of the range of possible future equity returns they can't possibly describe the range of future bond returns.
I will repeat again what has not yet been refuted or accepted by others: historically bonds have provided a useful and rewarding diversification in a balanced retail portfolio. However, at current (artificial yields) they can't possibly fulfil that role from this point. They should not be in a retail investment portfolio.

 

[Protocol]

https://www.frbsf.org/economic-research/files/wp2017-25.pdf

The Rate of Return on Everything.

Abstract
This paper answers fundamental questions that have preoccupied modern economic thought since the 18th century. What is the aggregate real rate of return in the economy? Is it higher than the growth rate of the economy and, if so, by how much? Is there a tendency for returns to fall in the long-run? Which particular assets have the highest long-run returns? We answer these questions on the basis of a new and comprehensive dataset for all major asset classes, including—for the first time—total returns to the largest, but oft ignored, component of household wealth, housing. The annual data on total returns for equity, housing, bonds, and bills cover 16 advanced economies from 1870 to 2015, and our new evidence reveals many new insights and puzzles.

Keywords: return on capital, interest rates, yields, dividends, rents, capital gains, risk premiums, household wealth, housing markets. JEL classification codes: D31, E44, E10, G10, G12, N10

 

[Gordon Gekko]

You thought wrong. That's the key takeaway from this thread.

I don’t think that’s accurate.

Equities “have out-performed all other asset classes over almost any reasonably long-term time frame”...we’ve seen a period where that hasn’t been the case...but note the use of the term “almost any”.

 

[Sarenco]

Equities “have out-performed all other asset classes over almost any reasonably long-term time frame”...we’ve seen a period where that hasn’t been the case...but note the use of the term “almost any”.

I would have thought that 20 years was a "reasonably long-time frame".

 

[Sarenco]

They might do 6% p.a. or they might go sideways.

Or they might fall.

Japan had the largest stock market in the world by market cap back in 1989. Japanese stocks have still to reach previous highs reached over 30 years ago.

The US stock market took 25 years to fully recover from the 1929 crash.

We simply don't know whether or not stocks will outperform bonds over the next 20 years.

I will repeat again what has not yet been refuted or accepted by others: historically bonds have provided a useful and rewarding diversification in a balanced retail portfolio. However, at current (artificial yields) they can't possibly fulfil that role from this point.

Bond yields fell (from historically low levels) during the most recent stock market correction. I can't see any basis for your assertion that bonds cannot diversify an equity portfolio going forward.

 

[Fella]

I don't worry so much what my actual return is , that is totally out of my control . Of course I would like it to be positive .

But I try to make the correct decisions.

When I choose investing in the stock market I make this choice because I believe it has he highest expected return .

Of course we only have data going back since the beginning of the stock market from what I can find is that bonds averaged 2.5% over 145 years V's stocks 6.9% .

I would expect stocks to out perform bonds over my lifetime investing but it may not but over multiple lifetimes people should expect a higher % returns from stocks . Of course each person only gets a limited time frame to invest in so it's down to chance , but your expected return is higher with stocks.

 

[Gordon Gekko]

I would have thought that 20 years was a "reasonably long-time frame".

No; that’s not what’s being said. Finding one 20 year period where that has not been the case does not disprove the fact that equities have outperformed over almost any reasonably long time frame. It’s essentially citing the proverbial needle in the haystack as something other than merely that.

 

[elacsaplau]

If we compare the 20 year return of the S&P against the US Aggregate Bond Index for the last 300 month ends (i.e. the last 25 years), there has only been 7 months where the Bond Index outperformed.

 

[Sarenco]

@Gordon Gekko

If you re-read the OP you will see that the outperformance of long-term US treasuries over a 20 year period is not without precedent.

And 20 years is not the longest period where long-term US treasuries outperformed US stocks (S&P500).

And that's just the US.

Long-term government bonds have outperformed domestic stocks for periods of 20 years or more in every major industrial nation.

In any event, I read the phrase "almost any reasonable long-term period" as referring to the length of the period - not the frequency of such periods.

For the avoidance of doubt, I am not suggesting that bonds have frequently outperformed stocks over long time periods - they clearly haven't. Over the vast majority of long-terms holding periods, stocks have outperformed bonds.

Nor am I making any prediction - implicit or otherwise - about the future performance of any asset class.