Assessing and inferring from financials

[Dave Fagundes]

Last couple offers I've had out I made because the deals seemed good based on assumptions I made including 5% each for cap ex and repairs. These assumptions were ones I got from watching some of the BP.com webinars about how to use the pricing utilities on the site. 

In both cases, when I got the actual financials, the amount the owners had spent on each property was far greater than 10% of the NOI. In one case the higher numbers were largely due to what seemed to be some serious plumbing issues, plus replacing HVAC in two of the units. In the other case, they were due to roof replacement and a pretty major remodel of one of the units. But even backing out those major costs, the amount of cap ex/repairs were way higher than I'd predicted.

I walked away from both deals. That was the easy part. The harder part is what lessons to learn from these experiences. One possible takeaway is that I'm just going way low on my capex/repair cost estimates. One relevant fact is that both of these properties were older (1938 and 1961), which has to come with higher costs. One idea I'm thinking about is adjusting capex/repair estimates to the age of property, say 5% for each category for anything built 1990 or after, then 1 additional percentage point for each decade before that. 

 Another issue is what to infer from these about future expenses. My thinking is that if the properties cost X amount in a previous year, that meant that they were likely to cost at least X in future years, and probably more as they aged. In one case, the owner tried to convince me that the higher cap ex/repair costs were due to his getting the place ready for the market, and that I should expect lower than average costs going forward because he'd already done the heavy lifting. That is possible, but sure didn't seem likely. 

Would be interested to hear what the experts think of each of these: (1) what's the best way to accurately estimate cap ex/repairs so you can figure out if a deal warrants making an offer in the first place, and (2) any reason to think past financials do not reasonably well predict future costs? Thanks for any thoughts on this.