I have the following situation. I am contractually obligated to pay $100M, adjusted for inflation, annualy for the next 30 years to my counterparty. It would work like this? Today is t=0, next year t=1, etc.t=o. PMT = $100Mt=1. PMT = $100M * (1+ change in CPI in period year 1)t=2. PMT =PMT in Year 1 * (1+CPI in year 2).I think its farily straightforward how this works.I want a way to structure a program to hedge against inflation. 1) It should be possible to use TIPS? If so how would this specifically be structured. I have cash on hand today that can be invested in TIPS. 2) Futures? I understand these is a futures market based on CPI. How would a hedging program like this be structured.I need to be able to model how this would work and specifics would be much appreciated.Much thanks.

The best thing you could do is see if you could buy inflation indexed annuity bonds (that is what your obligation is), that would close to match the inflation risk. Secondly, you could buy TIPS contracts to get the right inflation sensitivity at various points in the curve, and then borrow fixed rate debt to get the interest rate exposure right.Finally, you could hedge using a real world exposure. For example, if you bought property, or a business whose dividends went up roughly in line with inflation.

You have promised the coupon payments on a TIPS bond. If you buy a TIPS bond to hedge, you're left with the return of accreted principal. If you can buy TIPS bonds in any maturity, you start with a 30-year one whose last payment exactly matches the accreted $100 million. You subtract the coupon from that from the 29-year flow, and buy a 29-year TIPS to hedge the difference. You continue that all the way back to zero.It is rare that I disagree with gjlipman, but the idea of buying an asset is something that would occur only to an economist. Asset prices do not correlate with the CPI. Over the past few years, for example, US dollar inflation has been very low, less than 2.5%, yet the dollar has plunged in value against gold, real estate, foreign currencies, oil and just about everything else.The problem is CPI measures out-of-pocket living costs, which economists think should relate to asset values, but in fact does not. If the value of real estate goes up, investors are willing to take a lower cash return on their investment, so rents can go down. The CPI measures rent, not imputed rental value of owner-occupied housing, or cost of purchasing real estate. If the government raises sales taxes, it counts as inflation, but if it raises income taxes, it does not; yet these will have (to a first approximation) the same effect on asset prices.

"only to an economist" - I've never been so offended! ;-)Yes, I agree that you shouldn't assume that any given asset's cash flows will move with CPI. But I suspect some move more closely than others (and a very few - say those arriving from government CPI indexed payments, match perfectly). And if it is 30 years, you'll want to make sure you know whether you are hedging the cash flows, or the value - those might not be the same (eg if there are different credit spreads).

OK. I understand. You are describing an immunization approach. For the moment, assuming all maturities are available, for the 30 year bond, I would start off with a bond such that the amount paid would be equal to the liability in year 30: This liability is equal to the PMT * (1+ CPI30). Solving for the investment...PMT * (1 + CP_I30) = INVESTMENT * (1+CPI_30) + [INVESTMENT * (1+CPI_30)] * coupon%_30==> Investment = PMT / (1 + coupon%_30)So in my case, I would need to invest $100M/(1+ Coupon%_30). So assuming the real coupon on this 30 yr is 2%...30 yr: Purchase $98.02 Million in TIPS for the 29 year bond:The Liability is now PMT * (1+ CPI_29) - (30 yr-coupon)Here is where I run into problems....PMT * (1+ CPI_29) - (30 yr coupon) = INVESTMENT * (1+CPI_29) +[INVESTMENT *(1+CPI_29)]*Coupon_29I don't see how I can solve this????

It's cash flow matching rather than immunization, although of course, a cash flow match is immunized.Where you have 1 + CPI_30, it should be just CPI_30 (setting CPI_0 = 1 for convenience). So:PMT*CPI_30 = INV_30*CPI_30*(1 + COUP_30)PMT is constant and known, CPI_30 is unknown today but can be divided out from both sides, COUP_30 is known today from TIPS quotes. So we can solve for INV_30 as:PMT/(1 + COUP_30)Which is what you have. This bond will pay:PMT*CPI_29*COUP_30/(1 + COUP_30)at year 29. You need PMT*CPI_29. So you need:PMT*CPI_29*[1 - COUP_30/(1 + COUP_30)] = INV_29*CPI_29*(1 + COUP_29)So:INV_29 = PMT*[1 - COUP_30/(1 + COUP_30)]/(1 + COUP_29)and so on.

Thanks for your help. I see where I went wrong. So, in general terms, solving for hte required investment in t-matiruity TIPS:INV (t) = [PMT - (SUM (from t to T) (INV (t+1)*CPN(t+1))]/ (1 + CPN (t))now I have another question for you. I am LONG inflation now, as I need to make these payments of $100M grrossed up for inflation fo the next 30 years. If I implement this TIPS strategy, I will be SHORT inflation and also LONG nominal Treasuries (correct?). If I am only concerned with hedging the inflation, and not the financing for these payments, and still want to utilize this TIPS strategy for inflation hedging, what can I do? Essentially, I will be appropriated funds regularly to pay this obligation, so the financing will be taken care of, but not all up front. I only now want to hedge out the inflation, using TIPS.Thanks

You need to make a leveraged investment in TIPS. One way is to buy the TIPS and borrow the money, but I assume this isn't what you mean. The other is to find the longest TIPS you can and use a model or regression analysis to delta-hedge your inflation exposure. The exact hedge protects you exactly. A regression based hedge can fail if, for example, short-term inflation flares up but long-term inflation expectations decline.The more different maturity TIPS you include in your hedge, the better it will perform, but you'll have to buy more of the shorter bonds for the same hedge impact.

QuoteOriginally posted by: gjlipman"only to an economist" - I've never been so offended! ;-)Yes, I agree that you shouldn't assume that any given asset's cash flows will move with CPI. But I suspect some move more closely than others (and a very few - say those arriving from government CPI indexed payments, match perfectly).Reuven Brenner has an column today discussing this issue.

My dearest grand master Aaron Brown:Sorry for bothering you sir. (I am a newbie on this forum.)Inflation arbitrage: It's possible and become a huge business once you can overcome two obstacles:Obstacle 1: Completly accurate estimation of de facto inflation rate(s) (I think this is almost impossible; it's a heaadache since I know that all the inflation rates now being disclosed are forgery.)Obstacle 2: Transformation of the custom of living for the counterparty (I know that only the alien has this prerogative)Ouch!!! I got a headache!?!?"That's my prerogative..."TooNeat

QuoteOriginally posted by: AaronReuven Brenner has an column today discussing this issue.That article was very interesting, I hadn't quite appreciated the artificiality of the measure, or the inappropriateness of obsessing too closely to it.Maybe I should revise my comments to saying get assets that are explicitly linked to CPI, rather than just ones that 'tend' to go up with inflation.