Home arrow Storage Model
The Storage Model
"The winter/summer spread and volatilities have collapsed. How much should I pay for storage?" The WTM storage model will give you an accurate answer! Other models tend to make you overpay.

Why This Model?

The WTM natural gas storage model is designed to help traders optimally extract and hedge both intrinsic and extrinsic values each and every day; the model is not just a valuation model! The WTM model is the first to correctly bring together the two features that must be incorporated into any storage model if it is to be accurate and is to give optimal daily trading recommendations:

  1. The model must possess the correct methodology for capturing the complicated, follow-on American optionality of storage, and
  2. It must do so on a set of realistically modeled cash and forward price movements.
No other storage model on the market can do both of these simultaneously. The Monte Carlo simulations don't do (1), and the tree models don't do (2).

What This Means simply is that your current model could be telling you to pay 30 to 50 cents over fair value for the demand charge!

WTM’s Storage Optimization Software (SOStm)
"Itís winter, my lease ends in March, and forward prices are contango through next summer. When should I start withdrawing?" SOS will show you how to withdraw in a way to maximize your remaining extrinsic value!

As discussed just above, SOS by WTM is a breakthrough natural gas storage model and is the first to marry together the two critical features needed for storage modeling. This marriage not only provides more accurate valuations, but also allows the model to give more metrics for optimally managing storage: thresholds on cash-to-prompt spreads for optimal injection and withdrawal that day; superior dynamic hedges that protect extrinsic value; results that impart whether capturing intrinsic or extrinsic is more important on a given day. See our back testing results for how well this "marriage" works.
No other model on the market can accurately produce these metrics, if at all. These other models have been around for years, they haven't measurably changed over that time (only the "marriage" of the two critical features discussed above would allow for that), they are used primarily for compliance, and they are not used for trading. So if your needs are primarily to meet compliance goals, the other models on the market will be adequate to do so. And while SOS can be used for compliance, doing so as a primary purpose would be a great waste of its potential!
The SOS model is among the easiest of the storage models to operate. And with SOS, traders can now run the model every day, receive pertinent trading results from it, and act. Here are the metrics you get for managing storage:
  • Accurate storage values, both intrinsic and extrinsic, based on optimal storage trading
  • Threshold cash-to-prompt spreads for optimally injecting and withdrawing
  • Dynamic hedge positions in forward contracts that protect extrinsic, not just intrinsic
  • A table of results indicating which value is more important to pursue immediately, intrinsic or extrinsic, and which inventory levels have the highest optionality

Below, as an example, are some of these results as the model shows them.

VaR Calculation

The above results came from a Gulf region forward curve in early May 2011. The contract is on one-BCF, the cycle rate is 1.5 times per year, standard commodity and fuel charges apply, and the lease term remaining is 10 more months. The actual results show more dynamic hedge (delta) positions to the right (all the way to the lease end month) and more results by inventory level below (all the way to the lease's maximum capacity).

The above results show that the current inventory of 200,000 MMBtus has a remaining value, both intrinsic and extrinsic, of $1.578 per MMBtu (of which $1.325 is intrinsic). Subtracting out the weighted average cost of gas for that current inventory gives total performance to date. The results also show that it is optimal to inject (+8,330) for the current inventory, prices, volatilities and operating constraints. The model's hedging deltas are shown to the right, too; these deltas are designed for maximum extrinsic protection.

The threshold cash-to-prompt spreads are interpreted as follows: If cash is 11.7 cents over prompt (or less), the optimal physical trade is to buy and inject; at 21.2 cents or greater, withdrawing and selling becomes optimal. In between, doing neither is optimal. Thus, prices, volatilities, operational constraints, and the current inventory in this example are such that optimal trading is heavily skewed toward injections.

The table of results by inventory is best looked at by examining the "Recommended Trade (MMBtus)" over inventory levels. All else equal, this table typically shows injections as optimal for lower inventory levels, withdrawals as optimal for higher inventory levels, and doing neither for mid-range levels. This pattern is a mathematically proven result in cases involving simpler operational constraints. One can think of the levels in which no trading is done as the set of optimal inventories to be at and trade to. The levels in this set, whether near empty, full, or something in between, indicate which value is more important that day: intrinsic or extrinsic. So when you are in early September, you're 75 percent filled, but cash is trading at 15 cents above prompt, SOS will indicate whether it is better to withdraw to get the 15 cents or just keep filling to collect the rest of the winter-summer spread.

User Manual
White Paper
Back Testing Results

Copyright © 2004 - WTM Energy Software, LLC. All rights reserved. Terms and Privacy
Designed, Developed, and Optimized by Grimnes and Associates LLC.