On the species level expensive AND arbitrary displays signal ability to waste time/energy without dying, and act as a good indictor of fitness. On the individual level... it is arbitrary and without meaning, as are most behaviors.
I guess a 'signal' would be the process of incoming sensory stimuli creating a sparse and discrete representation in hippocampus that maps onto the statistical averages of one's evironement accumilated over time as overlapping representations in the cortex.
Now, it may seem that I'm mixing contexts here talking about species level selection and then explaining how displays are interpeted on the individual level. The anatomical and chemical landscape of the brain is encoded for on the species level, and the environmental stimuli, which mediate gene expression, of most of a species are the same.
But that's not concious perception of the signal. That requires the cortical abstractions/representation to propogate along the ventral and dorsal pathways of the posterior cortex to identify what objects are, and where they are respectively through a hierarchical process of pattern extraction using serially linked receptive fields in definite spatial patterns. This information becoming available to the prefrontal cortex is when the entity 'knows' about a signal.
So, to tie it all together, at a species level: Most known environments for life will likely not select for increasingly expensive (energy and lipid consumption-wise) computational mechanisms like faster metabolic rates, a bigger brain, an extra lipid layer to store incoming bursts of nutrients, etc. At least in environments wherein the organism has little means to change it's surroundings. When organisms DO have an ability to change their environment cooperation between individual organisms suddenly becomes energetically profitable.
A simple example/model and comparison: Because of the nature of fluid viscosity and how it's flow around objects changes with the size of the object (gist, see reynolds number), there is a causal relation between the size of free floating osmotrophs in the ocean, like algae, and the net profitability of cooperation between individual organisms. Small open ocean organisms display limited cooperation (things like, say, releasing cellular machinery into the open ocean to prepare nutrients for use) because the dimensionality (3D) of the space and diffusive transport of particles dropping off almost completely beyond 1 cell radius, algae cannot release enzymes and
expect to 'encounter' them or their products again. They are cut off from communication.
Now, bacteria on the otherhand (what animal life seems to have evolved from; after the bastards lost their thick cell walls an a membrane or two so they could undergo cell fusion, but I digress) living in the sediments are able to change their environment because of limited diffusion through the very viscous sediments and other hard trapping of molecules. They can release exoenzymes as mentioned above and reasonably expect to encounter the enzyme, and whatever it's interacting with, again. Therefore extracellular space is therefore useful for doing 'work'. This creates an environment then that is enriched in complex nutrient uptake and release dynamics due to the non-linear behavior of bacterial growth which exterts an environmental pressure on the organisms that can be 'solved' by inter-organism communication/displays. It's all downhill to complex multicellular organisms from there.