What it would take to "normalize" interest rate spreads?

I've spilled a lot of digital ink over the bond market sell-off recently -- and whether the worst is over. I recounted serious past market sell-offs carried, where some key areas of technical support lie, and more. Now, here's some more food for thought:

I analyzed the spread, or difference, between 2-year Treasury Note yields and 10-year Treasury Note yields going back to 1980. This "2-10 spread" is a key measure of the shape of the yield curve. When you have a positive spread, it means 10-year notes yield more than 2-year notes. It also means the yield curve is positively sloped (if you were to graph the yields on several Treasury bills, notes, and bonds of different maturities, starting with the shortest-term Treasuries at the far left of your screen, you'd get a line that slopes up and to the right.) When you have a negative spread, the yield curve is "inverted" -- 2-year notes yield more than 10-year notes.

Anyway, over the past 27 years, the average spread is +0.775%, or 78 basis points. A few minutes ago, the 2-year note was yielding 5.04% and the 10-year not was yielding 5.20%. That's a spread of +0.16%, or just 16 basis points. Now, this spread can increase in two ways:

1) All yields can fall, with short-term rates falling faster than long-term rates

OR

2) All yields can rise, with long-term rates rising faster than short-term rates.

The first method is generally "bullish" for the markets. The second method is generally "bearish." We're getting the second method -- in spades.

Now here's one last thing to consider -- to restore an "average" spread of 78 basis points ... even assuming yields on 2-year notes increased no more from here ... the yield on the 10-year note would have to rise from 5.20% to 5.82%. (5.82% new 10-year yield - 0.78% average spread = 5.04% current 2-year note yield)