Correlation of TIPS to inflation

[Chuckanut]

I got the following from Swedroe's column.

For example, from their first issuance in April 1997 through December 2011, the quarterly correlation of TIPS to inflation was +0.21.

Consider the correlation of Treasuries:
- For one-year Treasuries, the correlation was -0.18
- For five-year Treasuries, it was -0.37
- For 20-year Treasuries, it was -0.42


Can anybodyh explain in plain english how to interpret these correlation numbers?

[pete82]

yes, a 1-to-1 correlation implies price movements in sync (think the CPI and inflation).

the price of energy should have a pretty strong correlation to inflation (+.5, I made that # up but you get the picture).

price movements of TIPS are somewhat similar to the movements in inflation rate (+.21). Nominal Treasuries have weak or negative correlation to inflation. Nominal bonds are terrible hedges against inflation and rises in interest rates. they become more valuable during deflation and falling rates.

hope this helps.

[RetireBy90]

Normally correlation would be between 1 and -1. A correlation of 1 means they follow exactly, a correlation of -1 means that when one index goes up 10% the other goes down 10%.

So, a correlation of -0.18, -0.37 and -0.42 would mean that they don't follow each other, rather treasuries don't follow the inflation (cause of the Fed's actions?) but TIPS with a +.21 follow inflation much better than the regular treasuries.

[MichaelB]

Let me add to the confusion. Price correlations show how frequently two prices move in the same direction. They can range from plus one to minus one. Plus one means they both move in the same direction every time (100%). Minus one means they both move in opposite directions every time. Zero means there is no relation between the two prices. So a (-0.42) for 20 year treasuries is a pretty strong negative, meaning when inflation rises it is likely the treasuries will decline. Same for 5 year treasuries. The one year, at (-0.18) means there is a smaller chance this will happen, but it is still more likely than not to decline. Meantime, TIPs show a (+0.21) moderate propensity to increase in value when inflation rises. Given that TIPs are supposed to be the best protection against inflation, a correlation of (+0.21) is not a strong case.

These are quarterly measurements. They only tell you the direction of price change, not amount.

[panacea]

I don't think I can add anything to the wisdom these previous posters have provided. Just wanted to apply this concept to investment planning and express that it's this negative correlation that we look for for between asset classes when trying to add diversification to a portfolio. As most of us know, while two particular classes may be very risky on their own, the two together could result in a less volatile portfolio if they tend to move against each other. Negative correlation is harder to find than you might think with this global economy. Foreign and domestic markets are more correlated than they used to be... although emerging markets have less of a correlation than developed foreign markets to the US stock market. Stocks and treasuries/bonds remain some of the best classes to provide negative correlation, which is why we tend to focus on these two asset classes when we describe our portfolio (i.e. I'm a "60-40"). On the flip side, there's not a tremendous difference between small caps and large caps. With a correlation of maybe +.7 (my guess), both typically fall when the DOW or S&P is down for the day.

[Chuckanut]

Thank you to all. You have made these articles much more understandable.

[Gone4Good]

Let me clear some of the confusion.

Held to maturity, TIPS have perfect correlation with inflation as measured by CPI. At any point before maturity, the price is impacted by market forces and expectations for many things. This is no different than the correlation of monthly bond returns and their yield to maturity. A 10-yr bond with a 3% yield to maturity will only give you that return if you hold the bond for its entire life. Monthly, quarterly, and yearly returns on that bond will deviate widely from it's stated yield. It's just as accurate to say that a bond's return correlates poorly with its yield as it is to say TIPS correlate poorly with inflation. I think most people wouldn't see that as a very meaningful statement for bonds but feel it is very important for TIPS.

It's true that if you're trading TIPS, or hold long-duration TIPS, the day to day returns will not correlate strongly with inflation (remember too, that the CPI Index is updated monthly whereas TIPS prices change every second - high correlation between these two things shouldn't be expected). Held to maturity though, they're a perfect hedge to CPI.

[Midpack]

Quote:

Originally Posted by panacea I don't think I can add anything to the wisdom these previous posters have provided. Just wanted to apply this concept to investment planning and express that it's this negative correlation that we look for for between asset classes when trying to add diversification to a portfolio. As most of us know, while two particular classes may be very risky on their own, the two together could result in a less volatile portfolio if they tend to move against each other. Negative correlation is harder to find than you might think with this global economy. Foreign and domestic markets are more correlated than they used to be... although emerging markets have less of a correlation than developed foreign markets to the US stock market. Stocks and treasuries/bonds remain some of the best classes to provide negative correlation, which is why we tend to focus on these two asset classes when we describe our portfolio (i.e. I'm a "60-40"). On the flip side, there's not a tremendous difference between small caps and large caps. With a correlation of maybe +.7 (my guess), both typically fall when the DOW or S&P is down for the day.

Nice post. Being aware of correlations is a good way to check diversification/reduce volatility. But when there's a market meltdown (or more accurately a broad financial crisis), the correlations can break down temporarily at least and everything moves down together - as happened in 2008-09. Those are the times we find out what our individual risk tolerance really is...

[panacea]

Quote:

Originally Posted by Midpack Nice post. Being aware of correlations is a good way to check diversification/reduce volatility. But when there's a market meltdown (or more accurately a broad financial crisis), the correlations can break down temporarily at least and everything moves down together - as happened in 2008-09. Those are the times we find out what our individual risk tolerance really is...

So True!