I had a closer look at the facts and forecasts provided in yesterday's Westconnex Investor Presentation and, putting them together with the information available in other financial statements and presentations of the Company, I feel I have now a better sense of the possible/likely economics of this acquisition.
While I still see the deal as being somewhat unappealing, in terms of risk/reward relative to the existing assets, it is actually not as bad as I had initially thought.
I am still ambivalent as to whether I want to take up my entitlements or not, but in the mean time I thought others might find it helpful (especially given the scarcity of detail in the Westconnex Investor Presentation) to look at the way I modelled cashflows and arrived at an IRR for the deal.
I would like to emphasise that this exercise is for illustration purposes only, and it is just one possible way of measuring relative value in the context of an infrastructure asset portfolio.
My aim is to calculate, under reasonable Traffic Growth and CPI inflation assumptions, the expected Internal Rate of Return (IRR) of the deal, i.e. the flat rate that discounts all future EBITDA flows to the current Enterprise Value of the project (inclusive of future construction costs).
The Enterprise Value information is provided in slide 8 of the Westconnex Investor Presentation; as you can see, the total implied EV of the deal at the agreed purchase price (net of transaction costs) is 25.2bn$.
View attachment 1271135
In order to estimate all future annual EBITDA values (until the end of the concession), I will use the following starting points:
1) The FY18 Revenue information relative to the
M5 West Motorway, which is
already 50%-owned by Transurban, under an existing concession that will expire in December 2026 (and will subsequently be handed over to the Westconnex). This motorway generated 287m$ of Revenue in FY18 (see slide 40 of the FY18 Results Presentation), and its toll escalation follows plain CPI all the way to 2026 (see slide 65 of the FY18 Results Presentation).
2) The current market-implied level of Long-Term Australian Break-Even Inflation; this can be calculated from the prices of long-term nominal and inflation-linked Government bonds, and is currently around 2.30% pa.
3) The historical trend in traffic (Average Daily Trips) growth on Transurban-owned NSW and VIC assets, which is around 3.00% pa (see graph below).
View attachment 1271138
The expected Westconnex revenue breakdown by asset in FY28, as described by Management in slides 30 to 33 of yesterday’s Investor Presentation. Specifically, the expected relative toll revenue contribution in FY28 is:
M4 West / M4 East:
37%
M5 East / New M5 (East):
22%
M4-M5 Link / Rozelle Interchange:
16%
M5 West:
25%
First of all, I want to calculate the expected future Revenue values for the M5 West. Using 287m$ for FY18, and assuming a (1.030*1.023) annual growth factor (i.e. the compound rate of Traffic growth and plain CPI inflation) until 2026, for each year N I have:
Revenue(N) = 287m$ * (1.030*1.023) ^ (N-2018)
and, in particular:
Revenue(2028) = 287m$ * (1.030*1.023) ^ 10 = 484m$
Knowing that the relative weight of the M5 West in FY2028 is expected to be 25%, I can then calculate the expected Total Revenue from the Westconnex in FY28, i.e.
484m$ / 0.25 = 1,936m$ (actually 1,937m$, due to rounding).
With this information, I can now use the weights in point 4 above and calculate the expected FY28 revenue contribution for each Westconnex asset, i.e. (in bn$):
View attachment 1271141
We can now project these revenues per asset both into the future (i.e. between 2029 and 2060) and into the past (i.e. between 2027 and 2019), taking into account that the toll escalation rule for all the remaining assets (i.e. with the exception of the M5 West) is Max(4%,CPI).
One additional “boundary condition” that needs to be taken into account is the expected revenue contribution in FY19 from the M4 West, which is an existing observable asset that contributed pro-forma FY18 proportional revenue of 44m$ to the Transurban Group, i.e. 44m$ / 0.255 = 173m$ at the asset level, and 44m$ / 0.255 * 1.04 * 1.03 = 185m$ expected revenue in FY19. For the sake of simplicity, I will also assume equal contribution from the M4-M5 Link and from the Rozelle Interchange.
The overall result is summarised in the Revenue table below (the cells highlighted in green are half-year contributions):
View attachment 1271147
The revenue values are then projected all the way to 2060, with the condition that, starting from 2041, the toll escalation rule changes to plain CPI for
all Westconnex assets.
As a sanity check for the calculations performed so far, I notice that the bar chart in slide 11 of the Investor Presentation implies that the expected FY28 Revenue of the Westconnex (at 100% ownership) should correspond to 80%-85% of the expected FY28 Revenue of Transurban’s current Sydney assets (at 100% ownership).
View attachment 1271150
Using FY18 figures for the NSW assets, and projecting the corresponding revenues forward to 2028 (with 3.00% Traffic Growth and 2.30% CPI inflation), I see the previously calculated FY28 Westconnex revenue as amounting to
79% of the FY28 projected revenue for today’s Sydney concessions (at 100% ownership). That seems reasonably in line, given the likely differences in Traffic growth assumptions, and also in view of how approximate this bar chart really is.
From the Revenue Table above I can now calculate, for each year N, the corresponding EBITDA(N) as:
EBITDA(N) = [SUM(n=1,…,6) Revenue(n)] * [EBITDA Margin]
where n=1,…,6 are the six columns in the Revenue Table and [EBITDA Margin] is an assumption that can be made by looking at the current realised EBITDA margins for the Sydney-based Transurban assets. Looking at FY18 (see slide 43 of the FY18 Results Presentation), a flat EBITDA Margin of
80% looks like a reasonable assumption to me.
The Present Value of each EBITDA(N) can then be calculated as:
PV(N) = EBITDA(N) * (1+R) ^ (-N)
where R is a flat discount rate.
The IRR of the Westconnex project can finally be calculated as the value of R such that:
[SUM(N=2019,…,2060) PV(N)] = [Enterprise Value]
I see the Enterprise Value of 25.2bn$ disclosed by Management correspond to an IRR of
7.38% pa, which is higher than the 6.25%-6.75% range I had very roughly guesstimated in a previous post, but still significantly lower than the
8.14% pa IRR I see (using the same method and under the same CPI and Traffic Growth assumptions) for the
existing Transurban Group asset portfolio.
For a project of this size, involving a substantial amount of construction/development work, and with cashflows heavily back-loaded in the future, I would have rather expected a yield
premium to the current IRR, to compensate for the additional risk.
Instead, the effect of adding 25.5% of the expected future Westconnex cashflows to the existing asset portfolio, net of the dilution involved with the capital raising, is to reduce the overall Transurban Group IRR from
8.14% to
7.99% pa (calculated at the TERP of 11.85$/share).
As to the Group IRR implied by the Offer Price of 10.80$, I see
8.38% pa, under all the previous assumptions.
To conclude, from the information available and based on reasonable assumptions, I do
not see this Westconnex acquisition as adding any investment appeal to TCL. In fact, as it currently is, I no longer see TCL as a stock that can take up to 10% of my total investable capital. Accordingly, I am glad I halved my position on Thursday.
Having said that, at a TERP-implied IRR very close to 8% and with an embedded CPI indexation, I still see TCL as a solid real yield proposition; therefore, I feel reasonably comfortable holding a ~5% allocation at or around current price.
Hope this helps. As always IMHO, DYOR and all usual caveats apply.
(20min delay)