On Tuesday, February 19, I needed to adjust my butterfly trade. I wanted to do it by condor-rolling some ITM RUT call debit spreads. To explain, I had +10 MAR RUT 890/-10 MAR RUT 910 call debit spreads. To move those to +10 MAR RUT 900/-10 MAR RUT 920 debit spreads, I put in the following order:

-10 MAR RUT 890/+10 MAR RUT 900/+10 MAR RUT 910/-10 MAR RUT 920 calls.

My trade records show that I tried from 1:05 pm to 1:33 pm to get a fill on that trade, at one point being $0.12 under the mid on the credit I was asking. That's usually more than enough leeway to get a fill on the RUT, particularly when the order is directed to the ISE exchange, something my brokerage allows me to do.

That wasn't the only order I was working that day, with little success at getting fills on any of them. I like to condor roll spreads like that because the slippage is minimized when it's a single order. However, volume was low that day as traders trickled back after the long weekend, and important economic events were due later in the week.

My order was meant to raise the deltas on my big trade, something that was needed when the RUT was running up again. There just weren't enough traders out there who were willing to take the opposite side, taking my short deltas in exchange for anything anywhere at all near the mid.

When I back test trades, I tend to build in a little slippage, but I am not checking for volume or other considerations. Even if I build in a little slippage when I'm back testing, I "commit" the trade in my back testing system as if it has filled. Sometimes, in real life, you can't fill those trades as I was experiencing when I wanted to condor roll those debit spreads higher. Back testing has its benefits, but it's not always realistic.

A day later, the RUT had finally headed lower. That day, I could get fills, but did I really want to execute a lot of trades in a runaway market? Fortunately, my trade was structured so that it benefited from the move, at least on that Wednesday. If it hadn't, I might not have wanted to go all-out on a standard adjustment for my trade. For example, what if the trade had reached the point that I needed to move butterflies? Would I really want to do that in that fast-moving market environment? Of course, I would, if I'd been back testing. In real life, however, I know of several traders who took steps to stabilize their trades by flattening the deltas, but who waited for calmer conditions in which to perform more drastic adjustments. Fortunately, I wasn't in that position of needing to move butterflies or make other drastic adjustments on the 3+ standard deviation move that day.

So, what good is back testing, if it doesn't always replicate real-life trading? Back testing gives you a track record that inspires confidence in your trade. While all of this action was going on, many of us trading butterflies structured the way I trade mine were going through a discouraging time. Depending on the trader, we had either suffered lower-than-normal profits or endured actual losses for the previous two months. Because my plan doesn't allow for carrying my trades into expiration week, I closed those two previous months out at small losses while some of those who had limped into expiration week were able to eke out small profits. Butterfly trades can suddenly soar or fall apart in expiration week. It's just not part of my plan, so I was among those with small losses.

Moreover, I'd been struggling with my MAR trade for 26 days as these moves were developing. I'd managed to keep it profitable, but only if one squinted at the miniscule profit. Because I had thoroughly back tested the trade, however, I could look back at other periods when the trade had produced small profits or actual losses for a couple of cycles at a time and remember that, overall, the trade was profitable.

Back testing doesn't replicate live trading. It doesn't recreate all the emotions and wonky trading conditions, such as low-volume days when you just can't get a fill. What it does do is allow you to keep trading a workable trade during a period of drawdowns and know that, overall, the trade is profitable.

That's what keeps traders from abandoning their trading plans. It keeps them from not adjusting when trades need adjusting or over-adjusting when they don't need it, or jumping from trade to trade. That's what back testing does.

I have signed up for a proprietary trading system that isn't available for most traders, but you can find other subscription or fee-based back testing systems. Some brokerages may have them. I know people who use think-or-swim's "thinkBack" system, but I don't use it and so can't offer an assessment.

If you don't have back-testing capabilities, almost all brokerages allow you to reset certain parameters in your trade or a simulated trade setup, such as rolling implied volatilities up or down, changing the date or changing the price. Traders without back-testing capabilities could test their trades that way to determine how it might act theoretically in certain conditions. They could follow trades through on a simulated trader for several cycles, always pre-testing how certain adjustments might work and then following them through to determine if they worked as predicted. That might provide some confidence for live trading even if true back testing isn't available.

I'm so driven that I tend to get as emotionally involved with simulated trades as I do with live ones, but I know it's not the same for many people. Many feel a visceral shock when they move from simulated or back-tested trades to live trades. Having a track record with back testing can help prevent the paralyzing insecurity that sometimes accompanies that switch.