ALGO SPOTLIGHT #11: Convex Credit Spread Overnight

Nifty || Credit Spread || Overnight


What is it?

Imagine you’re driving through a mountain road.

Most drivers only pay attention to one thing — is the road going uphill or downhill. But an experienced driver notices something else too: how sharply the road is about to turn. That subtle change in the road’s shape often tells you what’s coming next, well before the turn itself arrives.

Convex Credit Spread Overnight works in a similar way.

Instead of only asking whether the market is moving up or down, it studies how option prices “bend” across different strike prices. When this bend stretches further than usual — a sign that pricing has moved away from its normal shape — the algo looks for opportunities where things are likely to snap back toward normal. It then builds a credit spread around that view and carries the position overnight.

You’re not trying to predict tomorrow’s headline. You’re reading the shape of today’s pricing and positioning for it to settle back into place.


How does a credit spread actually work?

A credit spread involves selling one option and buying another at a different strike, at the same time. The premium collected from the option sold is higher than what’s paid for the option bought, so the difference is pocketed upfront — that’s your credit.

The bought option acts as a safety net. Your maximum loss is capped and known before the trade even begins, which is what makes this a defined-risk strategy rather than a bet with open-ended downside.

Depending on which way the bend in pricing points, Convex Credit Spread Overnight builds one of two structures:

  • Bullish view → Credit Put Spread

  • Bearish view → Credit Call Spread

Simple analogy: Think of it like an insurer who only writes a policy when the premium being offered looks unusually generous relative to the actual risk — and who always buys reinsurance alongside it, so the worst case is known in advance. Convex does exactly this, just algorithmically, and it does it overnight, once a day.


Current Performance Snapshot

Metric Value
Since Inception Return 103.04%
6 Month Return 61.59%
3 Month Return 79.78%

These are strong numbers over a full cycle — but no systematic strategy moves in a straight line, and it’s important to look at the recent stretch honestly too.


Let’s talk about July

1 Month Return: -24.97%

Every systematic strategy experiences periods where prevailing market conditions are less favourable to its trading style. July was one such period for Convex Credit Spread Overnight. While short-term performance may fluctuate, the strategy is designed to be evaluated over complete market cycles rather than individual months.

A strategy built to read overstretched pricing will naturally have rougher months when the market doesn’t behave the way its underlying assumptions expect — for instance, during a sustained one-directional trend where the “bend” it identifies keeps stretching further instead of reverting. That’s not a flaw to hide; it’s a known characteristic of this style of trading, and it’s exactly why the next section matters.


When does it work best?

✓ Temporary overextensions in option pricing that tend to revert

✓ Stable option premium behaviour

✓ Normal volatility environments

When to be cautious

✗ Sudden news events

✗ Strong, sustained one-sided trends

✗ Extreme volatility spikes


Beyond Individual Algos: The Power of Combining Strategies

Imagine owning only an umbrella. It’s extremely useful when it rains. But on a cold, windy day, a jacket would serve you far better. Now imagine carrying both — no matter how the weather changes, you’re prepared for it.

Trading strategies behave the same way. Different algos perform differently across changing market environments, and no single strategy — however well designed — is built to lead in every kind of market. Rather than expecting one algo to outperform all the time, combining multiple, uncorrelated strategies can smooth out the overall investment journey.

Seeing it in the data

On Stratzy, you can check this for yourself using the Combine Algo feature — it lets you select any algo and see its correlation with others, along with a combined equity curve based on equal-weighted allocation.

Convex Credit Spread Overnight currently shows a correlation of just 0.03 with Damper Credit Spread and 0.11 with Curvature Credit Spread Overnight — both close to zero, meaning the three tend not to move in lockstep. Looking at a combined allocation across Convex, Damper, and Curvature over the past year:

The combined portfolio’s maximum drawdown was meaningfully shallower than Convex on its own, even while long-term returns held up well. That’s diversification doing what it’s meant to do — not eliminating drawdowns, but softening them.

With Stratzy’s Combine Algo feature, investors can build portfolios of multiple complementary strategies instead of relying on any single algo. It’s a tool worth exploring for anyone thinking about how their overall algo portfolio behaves across different market regimes — not just how one strategy performed last month.


Who is this for?

Someone who already understands the basics of credit spreads and is comfortable holding a defined-risk options position overnight, with exposure to gap risk between sessions. Given the drawdowns this style of strategy can see in unfavourable stretches, it tends to work best as one piece of a broader, diversified algo portfolio rather than a standalone bet.

Available on Stratzy. Check the past trades, review its correlation with other algos, and deploy it directly — no manual intervention needed once it’s live.

4 Likes

The actual DD of Convex Credit spread is 51% but not 32.5% as you shown. This issue is for all other algos as well. Please update the Combine Algo feature with correct DD numbers.

Hii Avinash,

As the team mentioned, this is under review and shall be optimised within few days.
Also, as requested if you could share this over email to support@stratzy.in along with your contact details then the team shall connect with you once this is optimised and keep you updated.

Hope this helps
Regards

1 Like

Hii @Avinash

This query of yours has now been optimised and we assure you that Stratzy takes feedbacks very seriously and works on them on immediate basis provided they are feasible.

Regards

Thank you the dd issue is fixed ,also show the Algo score for all the algos in that same page as previously shown, it will be really helpful.