Programming Myth? Pre vs. post increment and decrement speed

Question

So, I was marking a C++ test while our senior programmer was away and the very first question was something like this:

Without compiler optimizations, which method will run faster?

Method 1

for(int i = 0; i < 100000; i++) { }

Method 2

for(int i = 100000; i >= 0; --i) { }

Now, If you're like me and almost every other developer I've encountered (save now for one), you've been led to believe that pre-increment & pre-decrement operators (++i and --i) are inherently slower than post-increment & post-decrement operators. So naturally I went with Method 2.

However, our senior programmer returns and tells me that the two functions will run in equal time.

Thinking it was all some elaborate joke, I set out to prove him wrong and whipped this up:

#define ITERATIONS  (1000000000)
clock_t start = clock();
for(int64_t i = 0; i < ITERATIONS; i++)
{}
auto first_clock = clock() - start;

start = clock();
for(int64_t j = ITERATIONS; j > 0; --j)
{}
auto last_clock = clock() - start;

Much to my surprise:

first_clock = 25333
last_clock =  25277

For all intents and purposes, these methods are equal!

So, where did this myth come from? Was it true at one time and became false in newer compilers? Looking forward to an interesting discussion. :)