Suppose that an institutional investor had periodic obligations, such as a pension system that needed to make known cash flow payments each year for decades to come.
What do you think is a reasonable order to give to whoever is to manage assets to cover these payments?
No, not necessarily.
Think about an important objective rather than a limitation.
Absolutely.
That's the objective. Returns are nice, and limitations will come later. But for now, the focus is on making sure that you can pay the bills as they come due. This is a __liability-based mandate__: match the cash outflows with cash inflows.
No.
This could lead to risk levels that might really threaten the success of the firm.
So if you had projected payments to make in the next 10 years, what would you want to see from your investments in order to get this done?
Absolutely.
This suggests what's called a __cash flow matching__ approach. This is an immunization strategy, since the portfolio is "immunized" against market forces. If interest rates rise or fall, it doesn't really matter. You still have the coupons and face value payments all lined up to perfectly match what you will owe. So the inflows and outflows look like a beautiful reflection in a still pond:

Not really.
That implies a value play, but it also sounds too risky.
No.
That's not important here.
Now if you really wanted to set this up perfectly, what do you think might get to be a concern?
Yes.
There are transactions costs to consider. There's also default risk. But the idea is cash in, cash out in each period with nothing else to worry about. So there's no reinvestment risk since there's nothing to reinvest, and interest rates don't matter, either.
No.
Since the cash inflows match the outflows in each period, there's nothing to reinvest—no risk there.
Actually, no.
Interest rates don't really matter much with this strategy.
If you want fewer transactions costs, you could just set up the portfolio to have the same duration as the obligations. This immunization approach is conveniently called __duration matching__, just like it sounds. If your obligations have a duration of 8.3 years, then set up a portfolio with a duration of 8.3. Any shift up or down in the yield curve will be immunized.
Do you suppose this takes care of a yield curve steepening as well?
You're right.
No, actually it doesn't.
Parallel shifts are immunized with duration matching, but twisting and curvature changes can still cause problems. So this method is less perfect. Also, think about rebalancing. Which approach do you think might need more?
Yes!
No.
Duration matching is going to need more rebalancing.
Interest rates change, and that can cause the portfolio duration to change. So liquidity is a little more important with a duration matching strategy. With cash flow matching, you might think that there is no rebalancing to do, and that's fair. But there still might be times when the manager _wants_ to rebalance as various costs change. So it's at least a small consideration there.
Beyond these main two immunization strategies, there is a hybrid to consider called __contingent immunization__. Here, the idea is telling the manager, "Look, you can actively manage part of this portfolio, as long as the portfolio value stays above the present value of the obligations it's meant to cover." Then if it doesn't, what do you think would be a good idea?
Well, no.
That's riskier. Once it falls to the present value of obligations, fall back to full immunization.
Indeed.
Once it falls to the present value of obligations, fall back to full immunization.
This can mean cash flow matching or duration matching. Just no more active management since there's no buffer left.
Then you're back to hoping there are no defaults.
To summarize:
[[summary]]