What it is:
How it works (Example):
Let's assume Company XYZ issues $10 million of 10% coupon bonds that mature in 10 years. If after five years market rates on similar bonds fall to 5%, Company XYZ would be very tempted to redeem what are now bonds with a relatively high interest rate. To do this, it could borrow money at 5% and pay off the 10% bonds.
However, if the Company XYZ bonds have refunding protection, it can't. The refunding protection prevents Company XYZ from redeeming the bonds if the proceeds from the redemption are from cheaper bonds ranking equally or senior to the 10% bonds. So, issuing 5% bonds to pay off the 10% bonds won't work.
This of course does not mean that Company XYZ can never redeem the bonds early. It just can't do so with cheaper debt. If Company XYZ were to complete a stock offering, for example, or generate cash from another source, it would be free to redeem the 10% bonds as long as the bonds are callable.
Why it Matters:
Many investors confuse refunding protection with call protection. Call protection is much more absolute; it completely prohibits the issuer from redeeming the bonds before a certain time for any reason. Refunding protection, on the other hand, only prevents redemption in one circumstance. This is why a bond could have five or ten years of refunding protection but be immediately callable.
In a falling interest-rate environment, refunding protection can be particularly onerous to issuers and thus particularly advantageous to investors. The protection lowers investors' reinvestment risk -- that is the risk that the bonds will be redeemed and investors will have to reinvest the proceeds at a lower rate. In all other environments, the refunding protection doesn't change reinvestment risk or call risk exposure to investors.