Sunrun Inc (RUN) 2018 Q3 法說會逐字稿

Good day, ladies and gentlemen, and welcome to the Q3 2018 Sunrun Inc.

Earnings Conference Call.

(Operator Instructions) As a reminder, this conference is being recorded.

I would now like to introduce your host for today's conference, Patrick Jobin, Vice President of Finance and Investor Relations.

Sir, you may begin.

Thank you, operator, and thank you to those on the call for joining us today.

Before we begin, please note that certain remarks we will make on this conference call constitute forward-looking statements.

Although we believe these statements reflect our best judgment based on factors currently known to us, actual results may differ materially and adversely.

Please refer to the company's filings with the SEC for a more inclusive discussion of risks and other factors that may cause our actual results to differ from projections made in any forward-looking statements.

Please also note, these statements are being made as of today, and we disclaim any obligation to update or revise them.

On the call today are Lynn Jurich, Sunrun's Cofounder and CEO; Bob Komin, Sunrun's CFO; and Ed Fenster, Sunrun's Cofounder and Executive Chairman.

The presentation today will use slides, which are available on our website at investors.sunrun.com.

And now let me turn the call over to Lynn.

Thanks, Patrick.

We are pleased to share with you Sunrun's third quarter financial and operating results, along with progress against our strategic priorities.

In the third quarter, we added more than 13,000 customers, representing 100 megawatts of deployment.

This is the highest quarterly volume in the company's history.

We generated $86 million of net present value and created NPV per watt of $1 or over $7,700 per customer.

We are reiterating our full year guidance of 15% growth in deployment and growth in cash generation above this rate.

Demand is accelerating, and we expect Q4's growth rate to increase above 30% year-over-year.

This growth, plus investments in customer acquisitions and product innovation, will be achieved while delivering NPVs above $1 per watt for the full year.

Americans are united on the value of rooftop solar energy and batteries in their homes.

This especially resonates as we experience an increase in extreme weather events.

Just today, PG&E warned its customers of its decision to force 7 blackouts in up to 9 Northern California counties.

A centralized grid is not the solution to protect customers and provide reliable power when the wind blows.

Our Brightbox home solar and battery service gives people what they want, affordable, reliable power, which results in less pollution and healthier communities.

And not only can we provide product customers want today, but the assets we are deploying can be leveraged over time to help alleviate congestion on the grid when and where it is needed most.

Working collaboratively with utilities and grid operators, we can help America transition to a carbon-free and affordable power system.

This is the goal of our partnership with National Grid.

Renewable energy is a uniting force, and it makes me optimistic about the future of Sunrun and the country.

Two recent surveys underscore the popular view.

The first, commissioned by EEI, the utility industry group, found that 74% of Americans think we should use solar as much as possible.

And 70% agree that in the near future, we should produce 100% of our electricity from renewable energy sources.

The second survey from Consumer Reports highlights an overwhelming majority of Americans support greater reliance on cleaner energy sources, yet only 22% think that their utility is doing a good job investing in renewables.

It is not surprising that governments are responding with new renewable portfolio standards and other mandates that provide strong tailwinds to our business.

Sunrun's competitive advantages are accelerating.

As we continue to execute quarter after quarter and take share, we separate ourselves from the rest of the field.

This means we are more attractive to large national partners like big box retailers and homebuilders.

We have supply chain advantages that we are seeing play out today with batteries, and we are engaged with utilities to design the energy system of the future.

We're capitalizing on this market-leading position by investing in our direct customer acquisition and customer experience capabilities, which we believe will increase the moat around our business and deliver us to purer cost structure over time.

The results are encouraging.

Our direct business is growing over 50% year-over-year, an acceleration from last quarter, and we are increasing the adoption of Brightbox home solar and battery service.

We will have installed nearly 5,000 Brightbox systems by the end of the year and expect Brightbox installations to grow more quickly than solar only.

We have already launched the service in 8 markets, and it represents about 10% of our direct business overall and more than 25% in California.

Financial returns are attractive, and the service further differentiates Sunrun as the nation's leaders.

Brightbox provides customers with backup power, ability to manage time-of-use pricing and offers Sunrun additional revenue streams through grid services.

It also protects us against attempts by utilities to undermine the value of residential solar.

We have updated our agreement with Comcast to reflect learnings over the prior year and to accelerate the pace of the partnership.

Under the agreement, Comcast must spend an additional -- excuse me, an incremental $15 million in marketing prior to September 30, 2019, weighted towards our target market.

In addition, the partnership is now nonexclusive.

In exchange, we have reduced the minimum cliff for Comcast to vest into Sunrun common stock but have left unchanged the number of warrant shares issued per customer.

Comcast is delivering an attractive customer acquisition cost to Sunrun even with a continued rising stock price.

This amendment will make it easier for both parties to include results from the partnership in our 2019 plan.

Finally, we are cash flow positive even while investing in future growth and product leadership, and we have maintained our strong balance sheet.

I'll now turn the call over to Bob Komin, our CFO, to review Q3 performance and discuss that in some more detail.

Thanks, Lynn.

Customer NPV in the third quarter was approximately $7,700 or $1 per watt.

Year-to-date, in 2018, NPV per watt was $1.02, in line with our target levels despite the headwinds from tax reform and tariffs, along with the investments we're making to accelerate our direct business and product leadership.

Project value per customer was approximately $33,400 or $4.34 per watt in Q3.

As a reminder, project values is very sensitive to modest changes in geographic channel and tax equity fund mix, we expect project value will decline slightly over time, so with cost declining more, although on the short run, there can be quarterly fluctuations.

Turning now to creation cost on Slide 7. In Q3, total creation costs were approximately $25,600 per customer or $3.34 per watt.

Similar to project value, creation costs can fluctuate quarter-to-quarter.

Creation costs per watt were flat year-over-year.

We continue to expect creation costs to show modest declines for the full year even with the module tariff impact and as we continue to invest in growth in our direct business, as Lynn described.

As a reminder, our cost tax is not directly comparable to those of peers because of our channel partner business.

Blended installation cost per watt, which includes the cost of solar projects deployed by our channel partners as well as installation cost incurred for Sunrun-built systems, improved by $0.20 year-over-year to $2.52 per watt.

Install costs for systems built by Sunrun were $2.06 per watt.

We expect the adoption rate of home batteries to continue to increase, which will carry a higher per watt cost but also a higher project value.

In Q3, our sales and marketing costs were $0.73 per watt, about flat with our year-to-date results for the first half of 2018.

Our total sales and marketing unit costs are calculated by dividing costs in the period by total megawatts deployed.

Most of these expenses relate to our direct business, and these sales activities occur somewhat earlier than when the related systems are deployed.

During periods when we are growing direct sales rapidly, this timing impact causes reported unit sales and marketing costs to increase.

A higher mix of direct business also results in higher reported sales and marketing costs per watt, but it also means there will be lower blended installation costs per watt over time due to the higher mix of Sunrun managed installations at a lower reported cost per watt.

For the first 9 months of 2018 compared to a year ago, sales and marketing costs were higher, and blended installation costs were lower by approximately $0.20 per watt.

Our total GAAP sales and marketing expense increased 42% year-over-year as our volumes in the Sunrun direct business grew at a faster rate in last quarter and exceeded 50% year-over-year.

Sales cost per new customer in our direct business declined as we grew volumes at a faster rate than reported cost in the quarter.

In Q3, G&A costs were $0.23 per watt, a $0.04 or 15% improvement year-over-year.

Finally, when we calculate creation costs, we subtract the GAAP gross margin contribution realized from our platform services.

This includes our distribution, racking and lead generation businesses as well as solar systems we sell for cash or with a third-party loan.

We achieved platform services gross margin of $0.14 per watt, in line with recent trends.

In the third quarter, we deployed 100 megawatts, in line with our guidance.

While we don't manage the business for a specific mix between channel partner and direct, our direct business is growing at a strong rate and is a platform that enables Sunrun to be the desired partner for large national strategic and retail partners.

The direct business is also the platform focus -- the platform where we focus our initial Brightbox sales and installation efforts.

Severe weather and natural disasters did require us to spend operations for several days in 2 markets late into the quarter.

We always prioritize the safety of our installation crews.

So were not -- further hurricanes that affected southern -- South Carolina and Hawaii, deployments would've come in higher.

The increased frequency of extreme weather events, however, underscores the urgent need to address climate change now and to build a more resilient power system.

Our cash and third-party loan mix was 15% in Q3, also in line with recent levels and consistent with our outlook of low to mid-teens.

Turning now to our balance sheet.

Our liquidity position remains strong.

We ended Q3 with $275 million in total cash, a $5 million increase from last quarter.

We continue to forecast our cash generation will grow 15% or more, which would be $50 million or higher for 2018.

Quarterly cash generation can fluctuate due to the timing of project finance activities, but this represents our best view based on expected project finance activities for the remainder of this year.

We define cash generation as the change in our total cash less the change in recourse debt.

Also, please note that our cash generation outlook excludes any strategic opportunities beyond our current plans.

Moving on to guidance on Slide 9. We're reiterating our full year guidance of 15% growth in deployments and unit economics of $1 or greater per watt in NPV.

Now let me turn it over to Ed.

All right.

Thanks, Bob.

Today, I plan to address 3 topics: our strategy regarding interest rates; a recap of upcoming capital market transactions; and a review of net earning assets and capital run rate.

First, interest rates.

While the long-term interest rates that matter to Sunrun have nudged up slightly, the project finance market and returns for Sunrun remain healthy.

We remain on track to achieve our target of at least $50 million in 2018 cash generation and about $100 million in 2019 cash generation.

We're bullish on the overall finance opportunity and strategy for the coming years.

Our cost of capital is the sum of the benchmark 10-year risk-free interest rate and the credit spread that investors demand to finance our residential solar assets.

Over the last 3 years, the 10-year treasury has increased approximately 90 basis points.

Over the same time period, debt spreads for new residential solar assets have declined approximately 75 to 100 basis points, while investment grade advanced rates have actually increased.

Opportunity for further spread compression remains.

Although as spreads for residential solar senior debt are beginning to reflect the maturity of the asset class, spreads for residential project equity have not.

Spreads for residential solar project equity are still about 300 basis points, above typical utility scale financing cost.

Certain strategic investors are bidding utility scale portfolios even more aggressively, pushing many buyers out of the market.

Based on increasing interest in residential solar portfolio, we expect to see compression on the cost of residential project equity over the coming years.

For our existing systems, the majority of our floating rate debt is swapped for the weighted average life of the assets, significantly insulating us from changes in interest rates.

The value of these swaps is not lost in refinancing events.

For new assets, changes in capital cost do impact our cash proceeds.

For example, a 50 basis point increase in both our debt and project equity capital cost would reduce the proceeds we receive on new assets by approximately $0.12 a watt.

This can be offset by approximately 3% higher pricing or 4% lower cost or some mix of the 2. Currently, interest rate features imply a 5 basis point increase in the 10-year during 2019.

Finally, regulators typically set the rates incumbent utilities charge their customers in order to permit the utility to earn a return on deployed capital.

As such, retail electric rates are themselves sensitive interest rates, and we expect that over the longer term, higher interest rates will result in higher utility rates for customers, increasing our headroom to increase prices.

A multi-decade correlation between interest rates and utility rates demonstrates the strong relationship.

As we mentioned on prior calls, we are anticipating project finance activity will increase cash in the fourth quarter.

As before, we believe we will achieve the best possible execution by sequencing our transactions, first, in the public senior debt market; next, if applicable, in the subordinated debt market; and finally, to the extent desired in the project equity market.

We are well underway with these transactions and expect to launch the marketing of an asset-backed security transaction in the coming weeks.

We expect to accelerate the pace of ABS issuances next year.

On balance, the overall changes in debt market these last several years have been attractive despite rising underlying interest rates.

We've assumed rising interest rates in our own internal projections for this and next year and remain bullish as to our cash flow generation capabilities.

Moving to Slide 11, at quarter-end, net earning assets was $1.4 billion, an increase of $198 million or 17% year-over-year; and cash was $275 million, an increase of $39 million or 17% year-over-year.

Net earning assets is our way to describe the value of cash flow to Sunrun shareholders after payments to financing counterparties.

Turning finally to our pipeline, our debt and tax equity capital commitments already provide runway into next year.

And with that, I'll turn the call back over to Lynn.

Thanks, Ed.

Let's open the line for questions, please.

(Operator Instructions) Our first question comes from Michael Weinstein with Crédit Suisse.

Could you talk a little bit about the tax equity strategy in the quarter and the use of the lease pass-through structure?

How can we -- how should we think about the partnership flip versus the pass-through structure going forward at this point?

How much would that affect the results?

Sure.

So lease -- so this is Ed.

Great question.

So lease pass-through transaction that we've been deploying this year offers attractive overall proceeds to us, and so we chose to deploy it as part of our overall mix of capital for the year.

Predominantly, our capital deployed has been in the partnership-flip format, and it's likely to persist that way over the foreseeable period.

We just happen to have an investor offering and attractive capital cost in that format.

And given the proceeds were attractive to us, we executed that transaction opportunistically.

And as I mentioned -- sorry, maybe just to add in terms of the GAAP component.

At this point, we have about $54 million effectively of deferred revenue, which is reflected on the short-term lease pass-through financing line, which we expect will turn into net income as we complete the deployment of assets under that facility.

And I believe that is approximately -- will represent about $0.32 a share in fully diluted tax effect income.

Got you.

So in other words, if you wanted to take a look at the EPS, you'd add back the $0.32 a share for this quarter to make it more comparable to maybe how it might turn out in future quarters?

That would be sort of inception of the transaction to date.

In quarter, the number is approximately $17.8 million, which would be tax affected, and per share -- per fully diluted share, about $0.11.

Got you.

Okay.

And then looking into 2019, how do you think about the installation growth, especially since the Sunrun-built business grew 50%?

Would you be able to capture market share?

We like the acceleration we're seeing, and certainly, you see the Q4 guide would imply about a 30% year-over-year growth.

So it's really strong acceleration going into next year.

We're not in a position to give any formal '19 guidance right now.

But I would say, just big picture, we really like our competitive position.

I -- we think that on -- the advantages are increasing due to our market leadership.

And those are, primarily, I would say, due to we're becoming more and more attracted to the national partners like the big box retailers and the Comcasts of the world.

In addition, the product innovation we've been investing behind around Brightbox is becoming more critical as well.

So we're starting to see some differentiation there.

And finally, I would say that our investments in the customer experience start to pay dividends when now we're at 200,000-plus customers, and that existing customer base helps us to continue to grow and really acquire customers in our most affordable channel.

So I'm feeling strong about our competitive position but not -- we'll be in touch with the formal guidance next quarter.

Great.

And one last question.

The renegotiation of the Comcast Cable, that's simply because of you have a higher stock price now than originally been?

It's really -- we've -- the partnership has been successful, and we both wanted to make it more specific for '19, is really the spirit of it.

So they both wanted to be able to accelerate the earning into the warrant to try to earn into -- in the year, and we wanted incremental spend and spend in specific target markets.

So that combination worked out for both of us.

Also note that it's not an exclusive relationship anymore, so it's a partnership that offers very attractive acquisition dynamics and something that we are now in a place after learning over a year to have more specific targets around how we can perform.

Our next question comes from Brian Lee with Goldman Sachs.

Maybe just a follow up on the Comcast topic.

Can you, Lynn or Ed, maybe speak a little bit to what you learned over the past year that's prompting the changes?

Are you not going to appreciate that there was always a bit of an open-ended runway with the 40-month window?

And 30,000 to 60,000 customers announced, it's much more finite.

Looks like 6,000 to 8,000 in '19.

But is this more of a call for them to move to action?

Because I do feel like in -- over the course of the past year, maybe they had been slower to move than you had anticipated.

And so is this sort of a push for that?

Or is there something else that we should be reading into this?

You should not be reading anything more into it than we just learned how to make a more effective agreement for both parties.

So we just -- like you said, it was a long period of time as originally contemplated with, 30 months.

And so as we both entered into the 2019 planning season, we're both wanting to get more specific about where we're specifically going to invest, where we're going to invest the dollars.

And so the way we restructured it achieved cycle.

So it's really more about getting specific for 2019.

And I'll just note, it's the same -- the -- none of the terms in terms of the ratio of shares to customer acquired has changed in any dimension.

So the acquisition cost has not materially changed in any way.

Okay, okay.

Fair enough.

And then maybe switching gears here a little bit.

I might have missed this, but the Sunrun-built installation cost was up about $0.10 versus the last quarter and then also over $0.30 up on a year-on-year basis.

Can you speak to what's driving that?

And is this a new trend here we should expect?

Or does this step back down in 4Q?

So Brian, this is Bob.

Good question.

So there is several dynamics in there.

When you look at it year-over-year, one is, a year ago, I think we were at the low point for module costs in Q3.

And as you recall, we did some module purchasing coming into the year that was pre-tariff.

We've burned most of that off so we didn't get the full benefit of it in Q3, so module costs are kind of at the peak of what they've been in the last year in Q3 of this year.

As we look forward, the spot prices for modules are lower.

The tariff begins to step down at the beginning of next year.

So as we look into next year, we expect that the module pricing will come down.

The other couple of factors include batteries.

So in -- we have been successful in selling Brightbox, comes with higher battery cost, which is reflected in install cost.

That part of it is going to persist more as we go forward.

So there is some higher cost related to that, but it's offset by higher project value and how we price.

And then most recently, we've also had a little bit of more of a mix, and it's the same kind of dynamic where we also price up for it.

There has been some more high-efficiency panels, which are at a higher mix of them recently.

And with the tariff on those, that also is a step-up from what those would have cost a year ago, and those are higher-priced panels.

Okay.

That's helpful.

Maybe just a quick follow-up on that.

If you guys would have any hardware or equipment issues, not suggesting that you are, but if you had any of those, would those be covered under workmanship guarantees?

Or would you flip the incremental cost on that?

And have you seen that the dynamic that's impacted you on any part of your equipment supply chain, whether it be modules or batteries or inverters?

Do you mean workmanship -- you mean work through the channel?

Is that what you mean?

Yes.

Just -- if you had product failure issues that were above expectation, who is sort of on the hook for that?

And yes.

So Brian, this is Ed.

So there are a varieties of warranties that we collect in our business.

One is, obviously, product warranties.

We collect those directly from manufacturers, also indirectly.

And then we have third-party contractor systems from us.

We also have workmanship warranties from those third parties, and those workman warranties often also wrap the product warranties.

But the best defense, obviously, against any challenge is just good construction and careful selection of equipment, which is an area, I think, we've done very well at.

So we certainly haven't seen any sort of material deviation versus expectation in equipment performance.

But we do think that, that is in part the result of the fact that we focus carefully on good construction.

We audit and manage third parties carefully.

We audit and manage factories carefully.

And so a lot of that is really done on the front side, but we actually do have quite a belt-and-suspender degree of supports on the backside should anything arise in the future.

Okay.

No, that's helpful.

I appreciate that.

And then last one, I'll pass it on.

For you, Ed, on the ABS market, I think you had talked about doing something in 2019 previously.

I could be wrong there.

But does this new time line suggests you're seeing something in the ABS markets right now that's particularly attractive?

Or is this somewhat maybe preemptive to what potentially higher rates next year versus what you think you can get now?

Yes, sure.

Yes, so great question, Brian.

And maybe to clarify, historically, when we've spoken about accessing the ABS market, it's often been with a focus on aged assets, sort of those that are more than 5 years old, maybe post-flip assets.

We do still, as always, expect to do a transaction like that early next year.

What also has happened is that the terms in the ABS market for new assets has improved, making it relatively more attractive towards the capital as compared to, say, the bank -- syndicated commercial bank deals that we've done in the past.

And so the transactions that we are readying for the market currently is actually mostly a collection of new assets that have been very recently post-flip service.

Does that answer your question?

It does.

Our next question comes from Julien Dumoulin-Smith with Bank of America Merrill Lynch.

Can you hear me?

Hey, we can now, yes.

Hey, Julien.

Excellent.

All right.

Maybe to pick up right where Brian left off.

Can -- Ed, can you comment a little bit on the 2019 opportunity still just with respect to that -- those aged assets?

I think it was just about $1 billion of callable debt looking into next year.

And how you're thinking about that just, a, obviously, given the interest rate comments you made earlier; but also just in terms of the various financing choices in front of you.

And then also as well, if you can, I'll just ask the follow-up question now.

Capital allocation in 2019, I think I heard you mentioned some strategic decisions potentially, too.

Sure.

Well, let me answer the easy question first and then maybe just to clarify the last question if I understand it.

So yes, in the last call, we -- I talked about or having about $1 billion of debt.

That is on all asset, pre-flip and post-flip assets.

That's across the portfolio.

We did say that the opportunity existed to reduce the interest rate on that debt up to 100 basis points.

That opportunity still exists, and we've made some small progress towards it.

But expect most of that activity will occur in Q4 and in 2019.

On the -- your second question.

The capital allocation.

Yes.

I wasn't -- you needed to be more...

You've got -- you're accumulating cash now.

You've talked about this cash inflection.

What do you do with it?

Maybe said differently.

Got it.

Well, that's a good question.

So I -- so there are obviously a number of things we can do with the capital ranging from -- to the extent equity capital is necessary, safe harboring equipment against the step down in the investment tax credit, higher returns from that capital could include shareholder buybacks, accelerating gross dividends, sub mix of all of that.

I'd like to think that one of the things that we're recognized as being good at is allocators of capital.

And so we'll begin that process, particularly to assess the opportunity on safe harboring beginning early next year.

And out of the conclusions of that, we'll probably start to formulate a broader strategy around liquidity and cash.

Got it.

But you don't have clarity yet, maybe to come back to that point, about the extent that you need to put down for safe harbor?

That's still...

Yes, I think we've -- yes, we will be -- it's on the list of things to examine in 2019, what's the most efficient capital structure for that opportunity.

And just -- if I may just real quickly here, with respect to CCA opportunities, obviously, there's been some developments on that front in California.

Homebuilders, it's been a few months since some of the developments there.

Any progress potentially partnering with either one of those communities?

Yes, absolutely.

I would broaden it as well, too.

So on the -- just to answer homebuilders specifically, it's a segment we're very interested in.

We're in active discussions with the majority of the top 10 builders in California.

And what's becoming increasingly clear is that the third-party own model is the model that's preferred.

So I think we're really nicely positioned in that new homebuilding market particularly as it starts to ramp.

But again it's -- we think that ramp is closer to 2020 and really material volume.

So obviously, we're setting the foundation of the partnership now, but we really like our position for that market.

I think on the CCA, I would broadened that a bit to just a grid services work that we're doing generally.

And you heard in my prepared comments, one of the things we're very excited about Brightbox is that the Brightbox, you can monetize it twice.

You can monetize it with the customer, and there's real -- the customer is willing to pay for some of the value of storage because they like the background capability and the ability to protect themselves against the varying rate tariff.

But also, there is additional revenue stream available to -- in the ancillary services market, in the capacity markets, and we've done a lot of work on what is the total market size for that.

So if you look at just CapEx spending in the utility industry, it's $50 billion annual.

We -- with pretty punitive haircuts, we think the market available today for grid services from home batteries is $200 million growing to $3 billion by 2025.

So it is a market we're really excited about, and you'll recall that we did receive an $8 million investment from National Grid in order to pursue these specific opportunities within the next year.

So we have dozens of opportunities in the pipeline and enough certainly that National Grid became comfortable giving us $8 million to earn a return on projects that we bid just in the next 12-month period.

So again, those can take a few years to get going.

There's a long cycle on them, but we're really bullish about them as the real second wave of opportunity that we're going to capture as a business.

Our next question comes from Philip Shen with Roth Capital Partners.

This is Justin Clare on for Philip today.

So to start, just on the storage topic.

You indicated that you expect Brightbox installation to go faster than your solar-only installation.

I was wondering if you can just talk about how you see attach rates trending ahead.

And what do you think the attach rate could get to in 2019?

Just some color there would be helpful.

Yes, we're excited about the momentum in that.

We expect by the end of the year that we will have installed just under 5,000, and that is a doubling in the second half of the year versus the first half.

So there's no reason why that type of pacing can't continue.

We see, in California, last quarter, when we were in touch, the attach rate in California was 20%.

Now it's more than 25%.

So that gives you a sense for some of that pacing.

And I think if -- as the time the rate tariffs evolve, that will be a big factor in terms of what the percentage attached rate ends up playing out to be.

But our aim is -- where we think this industry goes is it's almost all Brightbox in 5 years.

Just throwing that out for the future.

I mean that's really where this is going to go, and so we're working very hard at innovating.

And we're having this success with today's current high prices, so we're seeing all this acceleration.

And we haven't -- because of the supply concerns, hadn't seen a lot of price decreases in the battery yet.

And there's so much innovation happening that we're optimistic that, that will come as well and drive another wave of adoption.

Okay.

That's helpful.

And then just on the supply constraints, is that an issue that you see at this point?

Or do you have sufficient supply to fuel the growth that you expect ahead of you?

We're well covered on supply.

As the -- the numbers I quoted are installation, too, so we're -- our installations are doubling back half.

It's not a bookings number, again supporting the fact that we're well positioned on supply.

It's one of the additional benefits of being the market leader.

And we're, as I said before, very excited about the innovation happening and the new entrants coming to market.

Great.

Okay.

And then maybe just one more for me.

So you're seeing really strong growth in the direct business.

It's been growing faster than the channel business for at least a couple of quarters here.

Can you talk about whether you see this trend continuing?

And then also, could you talk through just what you're seeing in terms of strength and weakness in different geographies at this point?

Sure.

The -- as always, we don't manage our business to mix.

The energy market is just a very local market.

It's a very dynamic market, and so we really look at it, bottoms ups, market by market with what is the most attractive way to acquire customers.

So right now, direct is growing very quickly.

Part of the reason is because we do have these strong national partnerships, because we are investing in the Brightbox product, and it's really where we're investing the most in the customer experience area of the business, which -- so that's what we're seeing now.

But we don't manage the mix, and we're really careful about not projecting to that either because that doesn't drive the right behavior we want.

The behavior we want to drive to is long-term competitive advantage and the NPV margin.

So we don't put targets out there because I think it would make the company make the wrong decision.

Our next question comes from Joseph Osha with JMP Securities.

So on storage -- most of my questions just got asked, but one quick one.

What roughly is the average sort of size for Brightbox that you're seeing going in?

Hey, Joe.

It's about -- maybe a little bit more than 10 kwh in the continental U.S. and maybe twice that off an island like Hawaii.

Okay.

Great.

Then moving on to this interesting transaction that you talked about, Ed.

A couple of questions there.

First, would it be fair to assume that we're talking kind of 100 or 150 megawatts in terms of the size of the package you're contemplating?

Would you care to comment on that?

Yes.

We obviously haven't announced formally the transaction.

Most of our transactions have been of a larger size, but not slightly.

But we'll be in touch in the coming weeks as that transaction begins to unfold.

Okay.

And then here is just more of a philosophical question.

Let's imagine that the market for new pre-flip assets continues to develop and be good, obviously, you have a bunch of seasoned assets that you can securitize over time as well.

But over time, could we see this evolve to a situation where your basic label to go out and create assets and almost immediately securitize them?

Could you see that being just the way the world looks in a couple of years?

It's close to how the world works today actually, Joe.

So we have to -- the nature of an asset-backed security transaction is you have to close it on a fully deployed set of assets.

So you do need to fill a certain number of tax equity funds, true them up.

The dust has to settle, and then you can term them out in the ABS market.

But that is supported in today's market, but you do have to warehouse them on your balance sheet for a little bit while you package them up to get them in a final format.

For sure.

But perhaps, I should have described it differently.

I knew we talked -- you talked about that other transaction next year that you're looking at doing with seasoned assets, and I'm just wondering if we could see a situation where the market for new assets is such that, that -- that really, that's where all the monetization occurs, and you're not out doing -- that doesn't end up requiring that you're out in the market doing transactions with seasoned assets, if you see what I mean.

So I think capital costs are likely to continue to be slightly lower for seasoned assets over a long period of time.

That said, certainly, the efficiency on the early assets has been improving, but advance rates in capital costs for both sets of assets have been improving.

So generally, just good news across the board.

But certainly, the most significant improvements have come on the new assets.

All right.

So the spreads are compressing faster there.

That was -- and then final question would be, do you all have any sort of particular preference in terms of buybacks versus dividends if we do get to the point where we're thinking about cash returns to shareholders?

I mean, I think we'll -- we were -- that would be something we would want to decide at the time.

Certainly, it's a board level decision.

But as we consider opportunities for cash including dividends, buybacks, other activities, those certainly are on the list.

Our next question comes from Colin Rusch with Oppenheimer.

Just thinking about your growth rate here in 2018 and as we look forward, an infrastructure that you have in place to grow, I mean, what sort of incremental investments you'd think you need to make to sustain that growth?

And does it make sense to start accelerating that a bit more as you start to see compelling economics and some more supportive financial instruments out there in the market?

Great question.

We're well positioned to support the type of growth rates we're seeing.

So again, we're very excited to come back and talk to you after Q4 when we forecast about the growth rate.

We'll step up to over 30% year-over-year.

So we think that acceleration will continue into the next year.

Q1 always have some seasonality, but we never like our strategic position more in where we sit today with the opportunities.

And the foundation is there, having -- we've been building that out for 11 years now.

So we have a very good sense for what we need to deliver these sort of high growth rates.

Okay.

And then just in terms of these ABS transactions.

Can you give a sense of size and cadence on those so that we just have a sense of how much of the portfolio is going to start flowing into these instruments?

Sure.

So obviously, that -- it's Ed.

That varies on market conditions, I think I've said on prior calls, which I think holds the case, is that it's probably the efficient transaction size is around $300 million to $400 million, like in that range somewhere.

And probably, it's efficient to not due to within a given quarter or something.

So you could expect to see that, I think, as much as that approximate cadence and that approximate dollar value.

But obviously, that's all subject to change based on market conditions.

This concludes today's Q&A session.

I would now like to turn the call back over to Lynn Jurich for closing remarks.

Thank you all very much.

We look forward to being back in touch soon, and we're excited for this growth acceleration into 2019 and the cash generation.

So we'll talk soon.

Ladies and gentlemen, thank you for participating in today's conference.

This does conclude the program.

You may all disconnect.

Everyone, have a great day.