Angelo,

You might have a look at this book:

The Intelligent Asset Allocator: How to Build Your Portfolio to Maximize Returns and Minimize Risk
by William J. Bernstein
Link: http://amzn.com/0071362363

or a similar one. Also, there are online resources that talk about Modern Portfolio Theory and the risk/reward trade-off.

If you have a portfolio that is maintained at 50/50 SPY/cash, then it'll put you at a specific point on the risk/reward curve. If you just park your money at 50/50 today, in 20 years, you could end up with 80/20, at an unacceptable risk/reward mix. You could end up with more money, but right when you want to cash out, another Great Recession could occur and you wouldn't be able to help your kids with college tuition. Of course, if your objective is to buy a Maserati when your mid-life crisis hits, then there's no point in diversification and rebalancing. Just gamble (since you'd be better off without the sports car anyway).

There must be a rigorous mathematical argument that if SPY is doing a random walk, then over time, you would not be able to gain by maintaining a 50/50 portfolio by regular rebalancing. Certainly, if you sell at the peaks and buy at the troughs, it'll be a win (as long as you consider trading costs). But, instead, you end up just making random decisions (albeit automatically, if you are running an algo), with respect to the market movements. So, the average gain is zero; you end up with 50% of the profit (or loss) compared to being 100% in the market. Of your initial capital, you've only deployed half of it, so by definition, you'll only have a 50% exposure (while you lose money to inflation by sitting on the cash).