By David Riester

What do I mean by "safe-harbor"? Spending money on equipment or labor earlier than called-for to ostensibly "lock in" something of value that threatens to "disappear". Being heavily exposed to subsidies & regulations, the renewable energy sector has often prompted one to consider safe harboring. Broadly, such moments have fallen into five categories:

1. Looming section 48 tax credit sunsets
2. Looming section 1603 "grant-in-lieu" sunsets (cash instead of tax credit)
3. Looming tariffs 
4. Supply chain shortages/problems (long lead times)
5. Inflation Reduction Act (IRA) related adders, such as "energy community" status

Here’s how it goes: A theoretical future thing scares everyone, threatening project margins/viability/timeline. Fear mixes with Action Bias – the instinct to do something instead of nothing. Herein lies a problem – there often IS something to do! and worse - it’s perceived (wrongly) to be simple. The action is to front-run the event by i) buying equipment or, ii) prematurely "starting construction", the usual means by which to safe-harbor.

The problem is twofold: First, the action rarely has the intended affect for the actor;  second, the reactive, panicky scramble to safe-harbor bollixes up the market.

Why doesn’t it "work"?

1. The feared event doesn’t come to pass, or isn’t as bad as expected. An ITC gets extended, a tariff is less severe than feared, those lead times turned out to be sandbagged, the legislation doesn’t pass. Pop quiz: do you know the original sunset date for the 30% solar ITC? December 31st, 2007. Do you know how many times the ITC was extended via legislation? Six and counting.

2. Creativity and resilience. The global renewable energy market sucks at many things, but we excel at troubleshooting, and we are survivors. Self-sabotaging tariffs emerge? Production shifts around southeast and central Asia faster than anyone imagines. 3 year MVA xformer lead time? Repurpose one from a defunct crypto op

3.     Stakeholder "Elasticity". Excel models tend to be myopic. If an ITC goes down, a module price up, a robotic model (most) inherently assumes the modeled project bears the full impact. In reality other stakeholders tend to take some of the hit. For more:

4. Inventory!

The investment (time, bandwidth, $$$, focus) usually has a bad ROI. 
The strategies are often pursued by less sophisticated folks on the periphery of the sector, because i) the problems can be understood without nuanced industry expertise, and the "play" – throw money at a problem, hire lawyers to CYA – is artless.

As we wade deeper into the waters of the IRA safe-harboring angles are emerging once more. We’d be well served to remember that sometimes the best thing to do is nothing at all.