Carmen Barone and the Barclays team graciously hosted me yesterday on the NYSE trading floor, and in the afternoon Marge Wywras at Knight Capital Group turned me loose with the traders on the Knight floor in Jersey City. That’s darned near a perfect business day to me.In reporting back to you on findings important to the IR chair, number one, floor operations play a crucial role still. These specialists, now called Designated Market Makers, are required to support issuers by actively placing bids and offers and committing their own financial resources. By wading into the stream of liquidity, 98% of which is automated on sophisticated mathematical systems, DMMs stabilize prices and help to foster orderly markets. Traders roam the floor with handheld devices controlling a variety of algorithmic options for working orders, which they deploy based on market conditions and customer preferences. They may follow the market, aim to affect a certain percentage of volume, aggressively push for alpha, and trade on a time-weighted or volume-weighted basis.

The floor revenue model has changed. Brokers are paid by the NYSE to provide liquidity, a form of rebate trading, which we’ve written about before. Just as grocery stores give out coupons to encourage customers to shop, rebates help exchanges and market centers attract liquidity from buyers and sellers, who then generate transactional revenue and data-services fees. The NYSE is supporting the DMMs by letting them collect fees that might otherwise go to super-fast market-neutral platforms, the high-frequency traders who own nothing and risk little in order to sit between buyers and sellers and churn shares back and forth.

The NYSE is stabilizing at some 25-30% share of traded volume now, down from the 85% (in listed issues) it enjoyed before the 2006 changes, but an improvement over recent pressure. The irony is that in an era of transparency, even the floor traders don’t know a great deal about the source and nature of the volume with which they interact. But they know a whole lot more than most. Our advice to issuers: make good use of their valuable access to market information.

My good friend Marge Wyrwas offered some eye-opening peeks behind the deceptively quiet façade of the building housing Knight Capital Group’s sleek trading operations. Wow. The floor is sliced up into sections providing sales trading, listed and Nasdaq equity trading strategies, ETF execution, bond trading, currency trading and even a new carbon trading operation. Billions of shares trade here every day. The degree with which Knight can knife through the maze and craze of equity trading and manage the execution of trades is really something to see. Two key takeaways: they do more real volume and less high-frequency trading than many think. And they blend the marvel of human insights with the magic of ultra-modern algo technology.

Personally, I was pleased to see again that we’re keeping right up with what’s happening. In fact, we’re tracking and clustering executed order flow to uniquely capture the nature and origin of liquidity in way that’s not being done, even at that trading level.

Why does all this matter? The IR role is an information role. Become the expert on market structure at your company and redefine the way IR is measured. It’s valuable to your management team, and cool to boot.

Switching gears, I owe a mea culpa. We’ve berated the exchanges for fueling conditions that constrain real investment – fragmentation, rebates, direct access, sponsored access, high-frequency trading, flash orders etc, et al, since data and transactions are keys to exchange prosperity. But Duncan Niederauer’s interview in the weekend Wall Street Journal (see link below) was the best call yet for return to capital formation in the equity markets. I am now cooking up a comfort-food casserole of crow in the crock pot.  I did drop a note to Mr. Niederauer saying so, too.

A word on the markets and then let’s talk about the size of trades and why they matter to your IR efforts.  These renewed swings of a percent or more lately on major measures reflect forms of statistical arbitrage in which systems tweak assets and seek small gains, while all around the noise plays. We do not see anywhere that more than 10% of volume reflects conventional value investing.

Which leads us to trade size. SEC, we hope you hear this. Issues of all sizes and shapes reflect nearly the same number of shares per trade.  Say your stock trades 2,000 times per day and your volume is 375,000 shares.  You average about 190 shares per trade. If you trade 5 million shares a day and 27,000 times daily, your average trade size is…about 190 shares.

What’s wrong with that?  We’re coming to it.  Trade sizes are a consequence of best execution rules, which require brokers to work within standard deviation. In time, there’s little deviation, or you get fined. We’ve defined a single entry standard for a vastly disparate market of 10,000 different companies.

What happens? The moment an institutional algorithm begins to make trades that consume liquidity – or “take” it – someone else’s algorithm looking for rebates leaps in between.  Thus we have markets where everyone must make and take liquidity, which becomes in time an end unto itself. Seldom – about one out of ten times – are active participants able to do more than tweak positions.

Plus, costs for doing so are low.  Advocates of this system trumpet low costs as though that’s the Holy Grail. The Holy Grail is a fat pipe of IPOs. Seen that lately?

In fact, the cost of capital formation is extremely high.  If you trade 2,000 times per day at 190 shares per trade, any value institution will be front-run before it ever gets close to 200 trades, or 10% of your daily volume. How will this value investor own your stock?

Likely, in baskets or ETFs.  It’s hard for an institution to buy or sell your stock as an end unto itself, unless it’s a hedge fund.

Now wait, there’s good news too. We’re here to make you look cool in the IR chair, after all.

If before, institutions needed 100,000 shares to realize gains, now they need only 10,000.  How? Suppose your prospective institutional investor buys 10,000 shares on a given day, and you trade 2,000 times per day. You may be nearly assured of appreciation in the coming 2-3 days, because the activity ripples through your market structure, changing the math and the behavior.

So, message accordingly!  Remind institutions that smaller commitments may produce appreciation – the desired outcome. Capital appreciation can be had for less money, so to speak. Hey, it’s an imperfect solution to a perplexing problem of rules – but we’re trying!

How will you measure these effects?  Market structure.  Learn it, love it, live it, IROs.  Rules aren’t going to change soon.  That’ll take time, and let’s all keep fighting that good fight.

First this on equity markets: Last week we noticed a surge in “volatility trading.” We’ve written before about these tactics that capitalize on volatility as the asset instead of the direction of the markets or a given security.

You should know about them because they’ll impact your price even if you trade less than 100,000 shares per day. We can find these things in market structure because certain firms specialize in it. When they show up, we watch what happens around them, thus discovering who else does it and what effect they have on market structure.

Under normal circumstances there wouldn’t be much. But we’ve woven such a tight risk-management net around algorithmic execution that everything reacts to everything else. Volatility traders who previously were using the VIX and similar measures perhaps have lately discovered that they can nudge risk-management algorithms around and produce volatility. It’s another unintended consequence of “systemic risk,” which can only exist when something artificially impedes natural failure.

Back to small caps. The average trade size in US markets today is less than 300 shares. We find in our pool that it’s closer to 200 shares. That’s partly due to the way the SEC measures “best execution,” or the quality of trades under SEC Rule 605. Broker-dealers must be within standard deviation of broad industry measures or they’re subject to fines. Naturally, over time executions become similar: about 186 shares per trade, regardless of market cap.

Say a multi-billion-dollar stock trades 26,000 times per day (and prices millions of times), while a stock with $500 million in market cap trades a thousand times. Both average 186 shares per trade.

A value institution is constrained by market structure from owning the smaller company. If they execute 100 trades, or 18,600 shares, high-frequency systems front-run and price them out of the market. Over time, the most efficient and compliant mechanism for owning small-caps are big risk-averse ETFs and algorithms cutting across hundreds or thousands of stocks, such as Renaissance Technologies or AXA or Dimensional Fund Advisors might run (through big prime brokers), or which Vanguard or Barclays iShares or Deutsche Bank Powershares might continually direct through their prime brokers and direct-market access channels.

Thus, small-caps become quant holdings. The problem is that small-caps and large caps alike need rational holders who stay the course through market swings and business cycles, the sort of thing that emotionless execution just won’t do.

This is a regulatory issue, IROs and execs. Structure is screening out your fundamental holders. How do we solve it? Speak up! Voice your opinion! Rules are supposed to create level playing fields, not advantage the dispassionate.