LBO Model: Entry Multiple Method - IRR goes UP with time?

Hey guys,

I have a modelling question, which is probably pretty basic but please excuse me.

I've built LBOs where we've looked at taking a public company private: start with share price, imply a share price with offer premium * NOSH = equity value + debt - cash + minority interest = enterprise value.

Uses = equity value + refinancing existing net debt (debt - cash) + fees
Sources: debt (in whatever structure) + equity as a plug (with or without management)

then flow through the model and get your return calculations.

My question is when it comes to valuing a private company and using a multiple for entry - I recently did this for my MBA program. We are looking at a secondary buyout of Agent Provocateur, and I believe the company should be valued at 11.8x FY2016E EBITDA (which I've estimated to be GBP11.6mm). So I do:

11.8x * 11.6mm = 136.9mm - debt (15.4mm) + cash (4.2mm) = 125.7mm.

Uses are: Purchase of equity (125.7) + refi net debt (11.2mm) + fees (2% of EV is 2.7mm)
Sources: debt (calculated as a multiple of FY2016E EBITDA) with equity as a plug.

The issue I have is when I calculate my returns. We assume transaction end of 2015A (March) so I do:

EBITDA for 2016, 2017 etc * 11.8 = implied enterprise value - debt + cash = implied equity value * sponsor equity (giving some to mgmt) = equity to sponsor.

My problem is that my IRR goes UP with time, which doesn't make sense. When I use an LTM EBITDA multiple everything works fine, when I make up a share price and # of shares as I would with a public company and imply an 11.8x FY2016E EBITDA multiple everything works fine, but its when I try to link it through to the 11.8 entry multiple the whole thing breaks down.

I've got some PE interviews coming up and the fact that this doesn't work is making me very comfortable. Any ideas would be greatly appreciated -- model is attached, nothing confidential here it was a school assignment.

Thank you all in advance for your time!